DC's 1756VK-assisted B4

FUB,

Those 2 pictures are a powerful argument.

But what about this block?



300K miles, head pulled for valve job, not blown gasket. Typical of what many of us just clean up and reuse. Guess I know the answer anyway.

I can't tell from the picture, but take a hard flat block and some 600 grit sandpaper and sand the deck just enough to see if there's some grooves in the deck to make your decision. My picture above was cleaned with a spinning yellow plastic fingers thingy so it will relatively easily go up/down in the grooves making them less noticeable.

A hard block will do a much better job of showing the indentations (or not) so you can make a more educated decision. You might be able to get some fine wires that you can use similar to a feeler gauge to measure the depth of the grooves (assuming you have them).

I had a cracked piston and signs of melting the lips on all 4 of them so if your engine has led a nicer life than mine had, you may not have much for grooves at all. Conversely, if you're not aiming for high torque/200+ Bar PCP's, a head gasket will last a lot longer in more marginal conditions as well.

You can see on the aluminum cylinder head, on the wide inner groove, the radial lines showing the head gasket moving in/out and wearing into the head. No coolant leaks that I noticed. This was on the factory assembled parts, FYI

FUB, I like the attention paid to the wear grooves from the head gasket's fire rings
Mine had them I'm fairly sure, but I paid them no real mind. Guess I'll see when it comes apart because it is leaking.

Digital Corpus said:

I'm really sure that this equated to significantly more than 125 ft*lb of torque on each bolt. I have a spare gasket and have a budget for head studs should this attempt fail.

Click to expand...

are these normal SHCS with a 12mm shank, or fully threaded, or shank reduced to minor diameter of thread?
If they're full diameter shank I'm interested to see how they hold up being that the working length is only from the start of the threads to somewhere around the 2nd thread in the block

[486] said:

are these normal SHCS with a 12mm shank, or fully threaded, or shank reduced to minor diameter of thread?
If they're full diameter shank I'm interested to see how they hold up being that the working length is only from the start of the threads to somewhere around the 2nd thread in the block

Click to expand...

How does the shank diameter/threading change the working length of the bolt?

Fix_Until_Broke said:

How does the shank diameter/threading change the working length of the bolt?

Click to expand...

stretch is concentrated at weakest diameter

Not having my block skimmed the first time around left me chasing head gasket issues - that and my block deformed...

skim it!

[486] said:

stretch is concentrated at weakest diameter

Click to expand...

Plastic deformation is, not elastic.

JFettig said:

Plastic deformation is, not elastic.

Click to expand...

Isn't the +180 deg when tightening the head bolts to get the bolts to stretch just past their yield point?
I think I was mostly commenting on using that torquing procedure with the standard shank SHCS.

Michael,

What does the repair schedule look like? Very interested in seeing the pictures of your block surface when you pull the head when compared to this picture:

http://forums.tdiclub.com/showpost.php?p=4922537&postcount=93

I did not check the block surface on the green car, just a quick check on the head surface. And that looked good.

The head has been installed and torqued down since I posted about it. A dynamic work schedule a and weather prevented me finishing things in the street. Rain yesterday prevented me from draining the oil and coolant and flushing it all. Just finished cleaning up injector bores and associated hardware. Oil will drain while I torque the injectors down a and pull the power steering pump to flush the coolant.

These are standard class 12.9 SHCS as assumed. Black oxide finish instead of blue die like I used before.


No warping detected on my end. If I was to precise enough, I won't likely know for a while now. Keep in mind I ran 12, 4th gear dynos on that head gasket. The leaks beforehand were attributed to rust & using a purely torque-based method of tightening.

Lurchalicious said:

Not having my block skimmed the first time around left me chasing head gasket issues - that and my block deformed...

skim it!

Click to expand...

If this fails again, I'll happily do so

[486] said:

Isn't the +180 deg when tightening the head bolts to get the bolts to stretch just past their yield point?
I think I was mostly commenting on using that torquing procedure with the standard shank SHCS.

Click to expand...

No, it has nothing to do with that actually. It's a much more precise way of torquing bolts to get a consistent clamping load. They frequently use that method for TTY fasteners because there isn't much margin of error.

JFettig said:

No, it has nothing to do with that actually. It's a much more precise way of torquing bolts to get a consistent clamping load. They frequently use that method for TTY fasteners because there isn't much margin of error.

Click to expand...

I've heard of it used in real big stuff, where you run the nuts down then slug over however many flats of the nut to get whatever amount of stretch, figured by the thread pitch.

Just thought the head bolts were TTY. Maybe they are, just higher grade bolts take more stretch to yield so these particular ones are no longer TTY.

Reading up on it some more, fastenal's got some relatively basic design information on bolted joints and in it they are saying the same thing I'd said earlier about the stretch being concentrated more toward the minor diameter of the threaded portion.

Bump bump bump!

I broken the turbo while I was doing the VE tests. That was almost a year and a half ago, if my memory is correct. I knew the failure had a limited number of modes so I wasn’t too worried and Life(tm) had been getting in the way. After disassembly last night the mode was revealed to be that of a dowel pin, whichhold the vane/nozzle adjustment ring’s bearings in place, had sheered itself. RyanP, aka Darkside is selling them Melett part, 1102-017-838 assembly and I’ve place an order for one.

Even though the dowels are 0.125” by a smidge under 0.5”, the vane’s tips felt up the turbine. I’ll replace the broken dowel temporarily until the new assembly arrives, but then I’ll be operational again.

No photos as of now since I’m migrating between computers and dealing with life, but I will take and share photos when I have new parts.

Alright, here we go... Photos first and then I'll talk.

All 11 vanes using macro lens on my cell phone. My computer array is still down for the count so sub-par images (for me) are the best I have atm.

What happens when pins like this:
Click to embiggen


Do this?
http://192.168.128.1/Misc-Pics/TDI-Pics/Turbo/GTB1756/Vanes/IMG_2486-low.jpg
Click to embiggen


This Happens!

Click to embiggen


Click to embiggen


Click to embiggen


Click to embiggen


Click to embiggen


Click to embiggen


Continued on next...

Click to embiggen


Click to embiggen


Click to embiggen


Click to embiggen


Click to embiggen



When I cleaned up the cage to the vanes, I also found these:
Click to embiggen


Click to embiggen



Vane Profile Comparison
Click to embiggen


Click to embiggen



The vanes were free floating and this image shows something of curiosity (image isn't rotating and I'll fix it at a later date)
Click to embiggen


One spot right from bottom center. That vane, which I confimed when removing the cage before photos, was flipped against exhaust flow, most likely causing the whistle I was heaving more of the time.

New Vanes!
Click to embiggen


Click to embiggen


New Cage!
Click to embiggen



So the post mortem is this: one of the dowel pins for the ring bearing sheered. THis occured when I was doing volumetric efficiency testing about 15 months prior to the repair. I've been busy, sue me . The inconel for the turbine is a much harder metal than the vanes. I do not know what alloy they use for the plate, vanes/nozzles/stators, shafts, control pivots (don't know if that is their proper name), bearing dowel pins, bearings, and control ring. The dowel pins are very lightly magnetic and, if the OE alloy is identical to the aftermarket alloy, becomes less magnetic after exposed to exhaust [temperatures]. But I digress, the loss of 1 of 3 bearings caused things to float. I lost another fixing/placement/alignment dowel pin that holds the slab of metal holding the vanes in one spot in the exhaust housing. I know I had a pin in there because the groove was bent to hell allowing axial play of the assembly.

I used some exhaust sealing cement/putty, sodium silicate, to affix and fill in some beginning play in the second pin which is press fit by the factory. The new vanes are 19.25 mm in length, fwiw. I've begun testing and adjusting my N75 map for a couple weeks and have made 4 iterative adjustments logging group 8 and 11. Why group 8? It give liters/hr for fuel consumption which is how the PID control registers P, I, D, D2 values. When I get to adjusting this, I'll post up some info, but with a strong and heavy disclaimer.

Anyhow, my B4 is back on the road and I'm loving all them torques

Pulled my Micro-1000 EGT gauge a couple of days ago due to two DMMs reporting it dead and shorted. Accidently installed it with 3 washers instead of 4 so the tip is about 1.3 mm past centering of the cross section. I realized that I never posted when I installed the thermocouple, where I installed it, and when I monitor it.

Most purpose built thermocouple data loggers report temps every second and this is slower than the original B4. The primary reason is due to the thermal mass of the thermalcouple which directly correlates with its time constant. You need 5 TCs to register an effective 100% impulse change. The Micro-1000 EGT probe I have measures at a 3.1 mm body with a 2 mm tip. This TC has no stated upper temperature limit. McMaster has plenty of thermocouples available with the fastest responding on being one with a 1/16" diameter body and tip. However, its max rating is 1690 °F, or about 920 °C. On the slower Micro-1000 I've registered over 1050 °C on it before on a WOT run.

Jeremy brought up a similar topic, here, a little while back.

Anyhow, my Micro-1000 still registers properly now that I've had it removed so I might put it back in. I've neutered my tune for a spell and limited it to ~24 PSI, but am using my torque limiter from the big kahuna. I'm slightly worried about melting my thin and stiny EGT with a 920 °C rating, but I've pushed 5+ seconds at temps over that, peaking at ~970 °C so far.

Additionally, I'm logging the millivolt output, doing my own cold junction compensation, and converting that back to temps to get a 4-5 samples per second readout on the new thermocouple. If it proves robust enough, I'll install one into my turbo outlet and a special port in the GTB1756VK's hotside to measure the temps about 1/4" from the vanes, which is where it matters the most. You can see the hold in the cage for this location in the last photo in my previous post.


Edit: Sample output. y-axis is in Celsius. Click to embiggen:

Alright, I am now able to align data together from EGT & VCDS logs. Took what I used to consider some fancy Excel work and is relatively simple now that I've been crunching data (30K-ish data points since they don't know how to use a database) in Excel for work. Yes, the plots coming up are "bling" but they aren't ostentatious, I hope. Anyhow, this leads to looking at operating conditions and parameters and fine tuning my tune.

Notes about the Plot
The logs used for the following were pulled for the sake of looking at injection timing, IQ limiters, and EGTs. I stopped off as a pseudo-rest stop on the way home from work, change the N108 adaption a tad, and proceeded to do a 1-2-3-4 gear pulls on flat ground. After finishing a stretch of road a little over a minute later, I have a nice, long, elevated freeway onramp that I did a 4th gear pull out to 4000 RPMs and drove home.


RPM / 100 = Black
IQ derived from Limiters = Orange
SOI Spec. & Act. = Grey & Blue
EGT = Yellow

The first three are plotted to the left y-axis and EGT is plotted to the right axis. The Excel file, aka log, is here (294 kib). You can change the time offset by modifying A44. EGT data copied from a separate file.

Click to embiggen:


Notes about the Tune
Timing map is the "standard" SOI map that stays BTDC from my tuner with a -1.9° adjustment applied globally. This was done intermittently due to troubleshooting my N75 map and is a stand in for greater modification due to the Diesel HPR fuel's Cetane rating of 75. Boost map is the modified/adjusted stock map from the 2.7l the GTB1756VK was fitted to. The N75 map is a custom one that I'd iteratively going through test-and-adjust on.

I have no way of verifying smoke output since the pull was at night, but given that I know I'm virtually a light haze to haze free during the day with similar test pulls, I'd expect the same. Boost is 2.2-2.3 bar absolute from 2500 to 4000 RPM in the tune, but with EGTs going up, I creep to 2.5-2.6 bar absolute.


Questions I have
Is it feasible to lower EGTs?
If so, will increased boost solve this even though I have no smoke?
What about SOI?
Should I open the vanes more via the N75?

I'm missing a PCP sensor (next year in Q2 maybe) and I'm on non-standard fuel. Opening vanes will decrease the likelihood of reaching requested boost (dependant upon PID control), increasing SOI increase PCP, increasing boost leans the AFR. All of these are acceptable methods to work by, but given that I'm at no smoke at ~20 PSI at "50 mg/str" I'm curious what action should be taken. Log the pump to see what voltage its at?

I'm looking to have EGTs peak at 950 °C/1742 °F, but then via "the tune," drop them to ~900 °C/1650 °F out to redline. Could I have my tuner do this? Yes, with a lot of trial and error and delayed back and forth and nit picking, but largely because I'm curious and I want to know how to do it myself.


Thanks for reading.
-DC

2 words: water injection (without MeOH).

Edit - elongated version: Trigger WI with a closed loop with injected quantity or EGT as the control input. If the latter, set the threshold EGT fairly low, say 1200F (calibrating with logged road testing as you've done, since there is substantial lag from the start of WI to when it has an effect on EGT. Always have a hysteresis hold so that the system is not "hunting" when near the switching threshold.

Pre-MAF?

Having AFR will help, especially if you can log it. The Innovate ones have a 0-5v output over a selectable range and are not that expensive.

The other thing that would be handy to have is exhaust pressures. I have not had the opportunity to play with a GTB1756 yet, but 17/22's once spooled you can let the vanes go wide open and it'll run just fine so I wouldn't rule out opening the vanes to help lower EGT's.

Some get too fixated on boost pressure control in the tune, probably because it's one of the only gauges most people have for feedback . You've got mass air flow, drivers wish (fuel) and an AFR target (smoke map) - boost is just a necessary evil to get enough MAF to keep the AFR in check. As long as you have enough MAF for the fuel delivered, don't worry what the boost is as long as it's under the max. When you add fuel, EGT's will go up and the turbo will spool and make more boost on it's own (this is what happened before VNT's ). I'm oversimplifying, but that's the general idea in my head.

See what the QA voltage is - if it's high then bigger nozzles will reduce EGT's (for the same fuel delivery after re-tuning). If it's low then you've got the option of advancing SOI. A few degrees will make a big difference in EGT. Start by retarding it 2 degrees and see how much higher/faster your EGT's go up .

Digital Corpus said:

Is it feasible to lower EGTs?

Click to expand...

Probably, you mention pulling two degrees of timing in there somewhere. I'm somewhere off the right edge of my SOI map, and adding only 2 degrees of timing in the top right couple boxes made my HX40 go from pretty responsive (and the pyro barrelling to the right with freightening speed) to just about as sluggish as you'd expect with that huge a turbine when bumped up to 15 degrees.
I don't have a good way for you to set your timing up well, it's pretty much guesswork without a dyno or a lot of datalogging at all load points. Ideally you set the timing for max power per unit fuel, in the real world you set it up like that, then pull it back until your head gasket survives.

Digital Corpus said:

Pre-MAF?

Click to expand...

god no, post MAF
pre compressor if you don't plan on winter, post intercooler if you do plan on winter.

Fix_Until_Broke said:

Having AFR will help, especially if you can log it. The Innovate ones have a 0-5v output over a selectable range and are not that expensive.

The other thing that would be handy to have is exhaust pressures. I have not had the opportunity to play with a GTB1756 yet, but 17/22's once spooled you can let the vanes go wide open and it'll run just fine so I wouldn't rule out opening the vanes to help lower EGT's.

Some get too fixated on boost pressure control in the tune, probably because it's one of the only gauges most people have for feedback . You've got mass air flow, drivers wish (fuel) and an AFR target (smoke map) - boost is just a necessary evil to get enough MAF to keep the AFR in check. As long as you have enough MAF for the fuel delivered, don't worry what the boost is as long as it's under the max. When you add fuel, EGT's will go up and the turbo will spool and make more boost on it's own (this is what happened before VNT's ). I'm oversimplifying, but that's the general idea in my head.

See what the QA voltage is - if it's high then bigger nozzles will reduce EGT's (for the same fuel delivery after re-tuning). If it's low then you've got the option of advancing SOI. A few degrees will make a big difference in EGT. Start by retarding it 2 degrees and see how much higher/faster your EGT's go up .

Click to expand...

AFR & EMP. Keep forgetting about those . Turbo speed is still on my list since a microphone isn't that environmentally robust. 5 minutes later with a few Google searches:

https://www.youtube.com/watch?v=ytcUmFx0BLg

About the MAF. I've been considering things like pitot tubes because it seems to unreasonably expensive and fiddly to fit a larger MAF housing. I need to re-do/add to/refine my VE measurements again. At least to 5K RPM. I have the math down for all of that already. Mark setup the tune as a hybrid so if the MAF dies or when I go over, it is still workable. Hell, I did 30 psi from 3K to 4K...

Running R520s and a 10 mm. If I'm fueling limited, I'm not adding a lift pump, but instead upping the line size to reduce restriction. Post-fuel-filter is TCO'd. I still instead to hybridize a 12mm plunger and head and have that sitting the the garage.

[486] said:

Probably, you mention pulling two degrees of timing in there somewhere. I'm somewhere off the right edge of my SOI map, and adding only 2 degrees of timing in the top right couple boxes made my HX40 go from pretty responsive (and the pyro barrelling to the right with freightening speed) to just about as sluggish as you'd expect with that huge a turbine when bumped up to 15 degrees.
I don't have a good way for you to set your timing up well, it's pretty much guesswork without a dyno or a lot of datalogging at all load points. Ideally you set the timing for max power per unit fuel, in the real world you set it up like that, then pull it back until your head gasket survives.

Click to expand...

The timing pull was from SOI. Cam timing is standard atm. I don't trust slow sensors and I know how to linearly compensate for a slow thermocouple, but they are still physical low pass filters. You're running standard diesel, yes?Also, if memory serves, you find it therapeutic working on vehicles and fabrication; I don't have that kind of passion behind that time.

Higher cetane ratings [can] cause a shorter ignition delay. Outside of the fact that environment temperature and pressure have a dramatic effect on this as well, a jump from 45-50 cetane to 75 can be an easy 1 ms reduction in ignition delay. (I need to find that paper again):


I consider myself an educating idiot, that is being constantly educated.

[486] said:

god no, post MAF
pre compressor if you don't plan on winter, post intercooler if you do plan on winter.

Click to expand...

Why???

Pre-MAF means the MAF registers the humidity difference, thus everything will still work, measure, and report correctly. Pre-MAF means water erosion via impingement on the compressor though. Our winters in SoCal leave me wearing shorts still. I'm not in the mountains, but can be up there in less than an hour.

I think I'll open up the vanes a few percent above 3500 RPM starting at an IQ of 41 (that's the current axis value) and see what happens. Log pump, MAP+MAF, and then go from there. Water injection is something I've expected to need at some point.

Maybe I'm just on the simple train lately, but is there an advantage of the ECU knowing about the water injection (pre MAF injection)? If you inject post intercooler, I'm guessing there won't be enough water injected to significantly change the actual mass airflow into the cylinder and this is just a high load conditions, therefore there's not much advantage to adding the risks you list above by injecting pre MAF.

Digital Corpus said:

Post-fuel-filter is TCO'd.

Click to expand...

What does this mean?

Digital Corpus said:

I don't trust slow sensors and I know how to linearly compensate for a slow thermocouple, but they are still physical low pass filters. You're running standard diesel, yes?Also, if memory serves, you find it therapeutic working on vehicles and fabrication; I don't have that kind of passion behind that time.

Click to expand...

My thermocouple is still a water heater thermocouple, so big fat and slow. Still haven't wrecked much, though I don't have a vane turbo any more.
Standard junk diesel, cheap as I can find with mandated 3% bio.
Stock cam and cam timing, never found compelling enough performance increases mentioned there to justify a cam or messing with the timing.
I do indeed do a lot of fabrication stuff as a hobby, less and less lately it seems. Really should find a new line of work, sucking me dry at the mechanic shop...

Digital Corpus said:

Pre-MAF means the MAF registers the humidity difference, thus everything will still work, measure, and report correctly. Pre-MAF means water erosion via impingement on the compressor though. Our winters in SoCal leave me wearing shorts still. I'm not in the mountains, but can be up there in less than an hour.

Click to expand...

Pretty sure the water would mess with the hot element MAF, as the water is never going to be evenly distributed until after the engine and it's all steam. Get a droplet of water on the element and suddenly it's reading full airflow.
Compressor erosion doesn't seem to be much of a factor with clean water, with dirty water it'll cause unhappiness, or with corrosive stuff like most alcohols. It also makes for even atomization and much better compressor efficiency. Since the air doesn't heat up as far with the water vaporizing in there it lets you get a higher density ratio for your pressure ratio. Think of it as an additional intercooler halfway through the compression stage.
Some of the water will condense out in the intercooler and if it gets real cold it'll freeze in there. Same as CCV gasses, really.

Fix_Until_Broke said:

What does this mean?

Click to expand...

I was intending to post photos but didn’t get to it before work today: I went from the stock 1/4” feed and 1/8” return to 3/8” feed and 1/4” return on the IP.

Fix_Until_Broke said:

Maybe I'm just on the simple train lately, but is there an advantage of the ECU knowing about the water injection (pre MAF injection)? If you inject post intercooler, I'm guessing there won't be enough water injected to significantly change the actual mass airflow into the cylinder and this is just a high load conditions, therefore there's not much advantage to adding the risks you list above by injecting pre MAF.

Click to expand...

"Wetter" air is less dense. Then again, I have a aluminum pipe from the turbo output to the rest of the stock intercooler piping. No reason to not plumb the water injection there, if I go down that route.

[486] said:

My thermocouple is still a water heater thermocouple, so big fat and slow. Still haven't wrecked much, though I don't have a vane turbo any more.
Standard junk diesel, cheap as I can find with mandated 3% bio.
Stock cam and cam timing, never found compelling enough performance increases mentioned there to justify a cam or messing with the timing.
I do indeed do a lot of fabrication stuff as a hobby, less and less lately it seems. Really should find a new line of work, sucking me dry at the mechanic shop...

Click to expand...

Similar reasons why I research the crap out of something before trying to implement it. I don't have the time to go ahead and dabble with the daily driver.

[486] said:

Pretty sure the water would mess with the hot element MAF, as the water is never going to be evenly distributed until after the engine and it's all steam. Get a droplet of water on the element and suddenly it's reading full airflow.
Compressor erosion doesn't seem to be much of a factor with clean water, with dirty water it'll cause unhappiness, or with corrosive stuff like most alcohols. It also makes for even atomization and much better compressor efficiency. Since the air doesn't heat up as far with the water vaporizing in there it lets you get a higher density ratio for your pressure ratio. Think of it as an additional intercooler halfway through the compression stage.
Some of the water will condense out in the intercooler and if it gets real cold it'll freeze in there. Same as CCV gasses, really.

Click to expand...

Water impingement and erosion for something spinning so fast will not matter if it is deionized, distilled, or tap water.

Yep. I followed this thread when it was active, back in the day.

It does, in a way, mimic a larger intercooler. I've avoided a FMIC due to the increased volume that contributes to turbo lag. I'm not racing or autocrossing so heatsoak isn't something I encounter much.

Custom measuring blocks work as expected.

I spent some time in Excel to make a pretty, indexed, human-readable lookup and comparison table. There are 9 fields that are unidentified, only one I haven't the slightest what it is supposed to do. The MSA15.5 software is setup for identifying 51 parameters in total before getting a buffer of 0's after it.

Also picked up a DAQ device that I can plug into my calculator or computer for $10; 4 analog channels at 2.5 ksps, 12 K samples of memory, non-realtime max sampling rate of 50 ksps. Need to make some circuits to tap existing hardware

Why is this interesting or useful? Well, As long as you know your driving altitude, temperature, and relative humidity, you can have a measuring block read out RPM, IAT (aka post-intercooler temperature), MAP/Boost actual, and MAF actual. Those latter parameters usually need 2 measuring blocks which then halve your sampling rate. Even though you're dealing with 8-bit numbers, you can oversample (take a lot of sample) that and get say 10, 12, or 14-bit depth and thus the sampling would take half of the time. Either average it all or add some convolution like Savitzky-Golay filtering (new favorite) to reduce quantization noise.

Again, why? Camshafts. I can more easily sample RPMs and make appropriate V.E. measurements.

I don't care that I'm rather late to do this for such an old platform, but it's fun. I'm slowly refinine my tune into something closer to "factory" level of precision. Thats ignoring the mistakes even they make like the A4 Jetta's hot start issue.

Identified the mystery voltage as a duplicate of the QA voltage. It’s identifier is different so I’m not sure if it’s inverted or what. The first mystery temperature is a duplicate of IAT and the other two reported as open. Also had an unknown bit output that might be for the auxiliary heaters, but since I’ve never been able to trigger those, it’ll be a while until I can verify. Also, there is a duplication of MAF Actual in the grouping. Now all I have to do is confirm the ADC chip and I can hack in EGT like what others did with the EDC15.

I did some review of the MSA15’s on Bosch’s website. Need to go through a second ~950 and pull those PNs & version numbers. The latter software/hardware revisions seem to be the literal transition to the MKIV ECU. I peaked at the measuring blocks of one for things like Measuring Block 13 where all 4 cylinders are indicated in the MKIV platform whereas differentials from Cyl. 3 are the only ones listed on more standard MSA15.5 tunes.

MSA15.7 ECUs started loosing the pressure port so I assume they had an integrated MAP/IAT like on the MKIV’s. Want to corroborate this when time permits.