Hi Jim,

Below are my comments to your post # 142:

===Originally Posted by pberku===
I am curious however as to why they found it necessary to change the design of the VFC housing, and its the cover. For a much smaller cost, they could have just substituted a lower temperature rated bms into the existing VFC design, but for some reason Mercedes concluded that they needed to do more than just change the bms.

===Jim's Reply:===
The term for what you describe is called a "design deficiency" and a team of engineers, electrical and mechanical, tackle this problem from a number of aspects to find a reliable and hopefully, low cost solution to the this deficiency.

When many changes are made, it's because the data from the many engineering tests, show that ALL of the improvements were needed. Companies don't spend money redesigning 'stuff' that's not needed.

===Jim, Here is my Response to your reply:===
Jim, Before a solution to a "design deficiency" can be resolved, the cause of the "deficiency" must be identified. In your opinion, and as per your web site, it seems that the design deficiency was limited to the bms. If so I totally disagree.

The bms is a device that acts on an INPUT, the input being hot ambient air. Based an that INPUT, It than presents an OUTPUT, the OUTPUT being its bending action. This bending causes the VFC's pin to release, which in turn engages the VFC. If there is no proper input to the bms, such as will occur if there is not adequate hot ambient air circulation around it, than the bms will not bend within the required time. Consequently it will not act on the VFC's pin. This in turn will cause the VFC to not engage, or to engage too late. So while your conclusions focuses on a bad bms design, my conclusion focuses on a bad ambient air flow design around the bms.
--------------------------------------------------------------------------------
===Originally Posted by pberku===
I am speculating that there was an airflow problem around the bms. Something in the original design may have prevented ambient air from adequately circulating around the bms, causing the bms to bend and engage the VFC at a much higher temperature than desired.
Phil

===Jim's Reply===
I believe I made that point a number of times in this post: and that testing in WATER is MUCH more condusive than testing in air especially in situ.

If the VFC/bms doesn't work in WATER, you can bet it needs to get much 'hotter' in air to engage. Now you are starting to see the light.

===Jim, Here is my Response to your reply:===
Testing it in water, as both of us did, [and as your remember we had different results], does not account for the fact that in real-life there could be limited ambient air circulation around the bms. (the INPUT) (see my reply above)

Lea also brings up a very valid point. Why did Mercedes not change the design of the same VFC for other models? The reason is likely that only subject engine had airflow restrictions around the bms, while other engine models, using the same VFC did not. This can easily be explained by different engine layouts characteristics, allowing for different ambiant air flow patterns around the bms.

As Lea also points out, it is not entirely clear that when Mercedes revised the VFC design, that they actually did change the bms. They did however revise the housing, likely to allow for more air flow around the bms, causing it to activate sooner.

So for now my conclusion remains that there was nothing wrong with the original bms design, but rather with the INPUT to it, ie: the limited ambient air flow around it.

Phil

Phil

Just a quick word to say YES. This is my opinion too. There's far too much concentration on the bms and I have been trying to point this out to Jim by questioning other aspects of operation. I think your comments are quite valid and true.


Jim

Please excuse my last rushed post. I must say my tone is providing food for thought not 'arguing' the point - viscosity for example. I'm merely offering my thoughts and looking beyond simply the bms.

I don't use the words copper/brass coloured because I cannot tell. Your's is very 'orange' compared to mine and so maybe brass. Mine looks quite dull in comparison I agree.

You have inappropriately changed what I have said and reposted it :( :(

I will reply latter.

Lea

At last, work is over. Now a little time for further comment:

LOL, yes, mine does have a subtle part number variation, 100% agreed. Also it 'looks' like it maybe of different metal construction, again I agree whole-heartedly and mine locks at 80C-85C water temperature. But, without taking measurement with TCs, I am unsure of the airflow temperature required to achieve this water temperature, due to a variation of temperature deltas around the cooling path and other factors including, rad water exit temperature and the averaging of the dash gauge/measuring circuit for example. Therefore I cannot comment further.

I'm simply trying to offer an alternative view as opposed to focussing on the BMS as we haven’t got the BMS data sheet and based on the assumption that MB and Sachs are actually in control of their BMS production and part numbering. I've tried to highlight this with MBs R129 119 production models/119 engines against dates coupled with the fact that Sachs do NOT define any ‘tropics’ BMSs for the R129 or W140 range – so how can I have a ‘tropics’ version if one doesn’t exist???


So:

How may the viscosity of silicone fluid effect the temperature of operation? Theory ONLY, but a possibility:

What I'm trying to emphasise is that the BMS alone doesn't have just two single states (bent/not bent), it bends progressively with temperature.


You say that under experimentation you cannot see it bend sufficiently (1.78mm) until 100C. I say it will bend from 80C to 100C but you may not be able to see this clearly underwater (refraction and other obvious effects) or when pulled out from the water due to cooling - the BMS has significant surface area and will cool quickly (and of course bend back) much faster than a digital camera can focus and take a picture. My comments and assumptions of the BMS bend maybe incorrect as I haven't carried out this experiment personally, so once I've repeated yours and Phil’s experiments I’ll will update.

For the purpose of this explanation I will assume the BMS bends 1mm at 90C, with this movement MAYBE the valve behind the pin allows (leaks) a minor volume of fluid into chamber 2. With an original clutch containing original silicone fluid this MAYNOT be enough to achieve lock. But now increase the viscosity of the fluid significantly (me adding something I think is about right) and with this same volume of leaked fluid MAYBE the clutch will lock.

I've mentioned before that mechanical engineering design isn't my forte so I will make no further comment on the internal operation of the clutch valve until I break mine apart and review :cool: :eek:

You will know that brass is an alloy (typically copper and Zinc) and the volume of added zinc gives its characteristic ‘yellow’ colour.

Yes, I meant ‘copper’ as in copper-colour; your site details several shots of BMSs and some clearly show a very orange looking BMS but some (Sac_2bms.jpg) dull. I will refrain from using metals to identify colours and I agree that mine doesn’t look anywhere near as ‘orange’ as yours (Sac_vfcfrnt.jpg as a reference) – maybe much less zinc? Maybe NOT brass at all?

All this said, there's still much supposition surrounding the bms as all could still have identical transfer functions. We REALLY need the data sheets on these parts – time for more Googling J

No, nor can I. But we know that MB does correct design issues, especially when there’s money involved – like all manufacturers.

Who told you it was an upgrade? (Menu 21). Search the EPC and you will find that 119 200 01 is only used on two engine models 119.974 & 975 EVER! Doesn’t sound like an upgrade to me!

Also, from ’89 through ’01 there has been VERY limited MB part number change to the VFC, most (bar one) are related to physical differences of the 974/975 engines - different number of fan blades so different fixing locations required on the clutch etc.

Also, Sachs do not list a ‘topical’ version for the R129 or W140 cars. So how can I have one????

Perhaps there is no problem with the R129 series, by using MBs own data system I cannot find any reference to VFC problems or coolant over temperature issues throughout the whole R129 life.

But I have a low temp cut-in VFC - maybe I'll attach a picture of my gauge :D ;)


Lea

OK some REALLY interesting info now...

I suspect the BMS is manufactured by KANTHAL - http://www.kanthal.com/

Follow the products section to 'Thermostatic Bimetal' and the data sheet for the TB1577 is there.

However, no information about the differing suffix 'A-GE' compared with '-GE'

So I'll email them now and we'll see if they reply. If not, I'll email from Work :-)

We'll get there soon.... :D :D :D :D I can feel it in my bones....

Now, I’d like to ‘do-a-LeaUK’ and summarize where we are. Note that I am only going to discuss LeaUK and JimF VFCs and their respective bms.

Abbrev: 119 200 00 22 = “00” ; 119 200 01 22 = “01”

JimF tests:
I have tested two (2) VFCs with “00” and “01” part numbers in water. Both results showed that they locked up around 100C. Both of these parts never locked up on the car in air . . ever! This was in their ‘stock’ configuration; they did lock with modification.

Have inspected many bms and have NEVER seen a bms that was marked different than the two that I tested. Yet, documents ‘claim’ that there are different bms in some of the various VFC assys. Still unresolved.

LeaUK tests:
He has an operational VFC on his car that locks up at yy C degs. I say ‘yy’ degrees because I think that the indicated temperature (85C) in the IC, may be low. But if the gauge is correct, then he definitely has a low temperature bms.

This conclusion stems from three (3) facts. 1) his bms has different markings (“1577” vs “1577A” for JimF); 2) it appears to be made of a different amalgam than the ones that I tested; 3) it actually locks up when mounted to the car.

Remaining Tests:
To prove that the LeaUK bms is different, it should be tested in water. It would be informative if a JimF bms could be tested at the same time. By doing this, you would see the different “bends” and these would be immediately visible.

Remember as shown in MENU#21, the two bms bent exactly the same. I suspended both of them ON THEIR SIDE with separation so you could see slight bends and could see different amounts of bend. No such thing happened; they bent the same. Based on the part number, “01”, it should have been different.

Testing Methods:
I’ve included this so as to point out that WATER is the preferred test medium because the properties and environment that the bms being tested can be easily controlled. We all should know this by now, but based on some posts, it seems to be missing.

A small vessel of water, big enough to contain a bms and/or VFC + bms is relatively easy to heat to a ‘homogeneous’ temperature. As such then the bms (or multiple bms) being tested will not be subjected to different temperature gradients is would be in “air”.

If testing of the LeaUK VFC on a MB car shows that it locks up at 85C, you can be assured that the bms in the VFC will show more bending than a JimF bms. So testing them side-by-side would make a ‘great’ picture and clearly show the different bends.

Also, you do NOT have to remove them from the water to take a picture if you use my technique; suspend the two in a small metal ‘pick’ that has small notches to keep each bms separate AND each on its side. Then take the picture from the top (straight down), the diffraction would not be a factor. And whatever it is, it would be the same for both.


My comment is ‘could-be’! But remember that the higher the viscosity of the ‘gel’, then the bigger the opening needs to be to get it to migrate. Little opening, high viscosity gel = little gel migration; little opening, low viscosity gel = more gel migration. Kind of a double edge sword.

BMS bend vs Pin Clutch engagement:
In one of the posts, you may recall that I modified a ‘stock’ pin clutch and removed about 0.030” from the end. It should be obvious now that I’m trying to see how much NEEDS to be removed, so that it will engage (lock-up) at “xx” C temperature. Trying to see what “xx” is versus pin clutch length. And make “xx” about 95C coolant temperature if possible.

So for a total pin clutch ‘travel’ of 0.080” or so, removing 0.030” inches has not helped; it has not locked up. The car’s coolant temperature showed 95C max.

Since the bms bends “a little bit” at 90C, that’s not enough bend to start lock-up. As I said before, the bms must bend quite a bit to engage the clutch; just a little bend is not enough to start gel migration. This is an almost new VFC w/ 'stock' gel.

Plans are to continue to test and hopefully the weather will get warmer, maybe even ‘hot’! If it still doesn’t lock, I will remove 0.010” more and then test it again.

Jim

A well written summary, summarising your thoughts; it was most needed and I see that you have actually agreed with me on a point ;)

I'm fascinated by your VFCs not locking on the car at all - there is definitely something wrong and I can understand your frustration especially as mine operates more than satisfactory. I wonder if your aux fans were to be disconnected the VFC would eventually lock at all?

Are there differences is air flow/circulation between the 00 and 01? I also wonder if MB changed something internally such that less travel is required. This would mean a lock at lower temperatures.

My coolant sensor is around 6 months old and I've checked it's output R against the meter's display. At 80C it's within measurable error.

BMS testing. I'm quite happy to send my BMS over to you for comparison (and I take your point about the refractive index - as long as the test is a comparison, it's negligible) when I receive my new clutch. This will provide for extra info on your site too :)

My only comment would be to consider that the fluid is also under significant 'pressure' from centrifugal force and so I suspect that this has some part to play too. Again, another horrible variable in the equation!

My summary so far.

1. My R129's VFC locks at 80C- 85C coolant temperature

2. I've refilled the clutch with 12,500CST silicone fluid and this may not be identical to the viscosity of the original fluid.

3. Due to the refill, the volume has certainly changed.

4. For the R129 or W140 chassis there is no 'tropical' VFC listed by Sachs

5. MB have NOT rolled out the suggested 'improved' E500/500E VFC MB PN: 119 200 01 22 to any other vehicle, EVER.

6. The BMS is probably manufactured by Kanthal (www.kanthal.com) - awaiting confirmation and data sheets.

7. The MB part number on the BMS doesn't necessarily reference the BMS only. It's the whole assembly.

8. I will have a new VFC shortly to test.

9. My BMS has a subtle part number suffix difference - a missing A

Lea

edit 8-05-05

10. My plastic clutch cowl is removed

It's not just me; read the many posts on MSF and you'll see that your VFC is the exception. So when you get your new VFC, you will be joining me and the others!

Not sure that you really will do this when you get the new VFC. You will need to use the old one almost guaranteed! But, yes by all means, I would test them as recommended. So I would welcome one that works!

The data sheets are on the web site; no need to request them. I've zipped them into the attached file fyi.

They "155" material (TB1577 cross reference) are in the ZIP file. Kanthal is part of the Sandvik Materials Group.

Jim

There is much reason to contact Kanthal as unless we can cross reference the 'A-GE' and 'GE' codes correctly the exact strips' data cannot be identified - there is no reference in the data sheets or web site.

Kanthal offer proprietary marking as a service and so MB/Sachs has probably used this. Also when considering the volumes purchased from automotive companies, suppliers bend-over backwards should any level of customisation be required :)

One thing I don't understand is why no-one has found this data or contacted Kanthar before? Maybe they have but Kanthar won't release the data, which would be surprising but not unheard off.

Once we have the A-GE and -GE parts correctly cross referenced, or the actual data sheets, there should be no need for comparative testing, although there's nothing wrong with carrying out our own experiments too.

BTW - I guess you also compared the chemical construction of the 1577 range as presented in the data sheets - Fe, Ni, Mn and C. No brass in sight! The predominant metals being Ni and Fe - these aren't 'orange' coloured - another point of confusion to add to the already interesting story......

Regarding sending my BMS over to you Jim, I trust you wont destroy it and if you can return it to me once complete I can't see there being any issue. Of course, if my new VFC operates as yours (and all the others you mention) I'll have to suffer the higher temps until testing is complete. My car failed to lock at all before I refilled it and who knows how long the previous owner carried on relying on the aux fans only! It'll be fine for a while. :cool:

Lea

Are you considering the entire system?
 

Hi guys,
After reading this long and interesting thread, I thought I'd contribute an unrelated but possibly useful incident to the mix. Our 1988 Astrovan's thermostatically controlled fan began running almost all the time, but the temperature gauge was not rising even above the cold range except on extremely hot days. I decided to check the thermostat first, as it was the easiest thing to get to, and was surprised to find that it wasn't completely closing due to a failure of the seat. Originally it had a plastic material that provided a seating surface for the valve, but the material had partially broken away, causing a gap of around 1/16" all around the valve. Upon replacing the thermostat I was amazed that now the fan came on for around 10 seconds when first started, then stayed off until either the engine was heavily loaded, as in long uphill pull, or unless the ac was on on a hot day. Also, the temperature gauge now reads correctly. I am assuming that the water flow from the engine through the center portion of the radiator was strong enough to heat the bimetallic strip excessively when the thermostat had failed. This brings to mind the thought that there might be differences in the water flow through the radiator on your various vehicles that could cause temperature differences in the air that passes by the fan's bimetallic strip. In other words, there might be more to the story than just the bimetallic strip.
Enjoying the discussion!

Regards,
Richard Wooldridge
1982 300D/4.3 V6

Hi Richard,

Your contribution to this thread is much appreciated. It will be nice if even more members were to contribute to this discussion. Your point that it could be a defective Thermostat is well taken, however in this case, Jim's claim is that the problem is wide spread, and not limited to his car only.

I have stayed in the background for the last few days, but have been reading intently the ongoing discussion between Lea, and Jim. As you know from my previous threads, Lea, and I, agree with you that there is more to it than the just the bms.

An close analogy would be a house Thermostat. If you mount the house thermostat in a location, where it can not properly react to changes in ambient temperatures, it will then not be able to turn the house furnace on, or off as designed. This does not mean that there is something wrong with the Thermostat, It just means that it was mounted improperly.

The same goes for the bms. If it can not detect changes in the ambient temperatures properly, because of inadequate air flow around it, than similarly, to the house thermostat analogy, it does not mean that there is something wrong with its design.

The solution would be to increase the air flow around the bms to allow it to react quicker to changes in temperature, and as per Jim's attachment in post #138, this is exactly what Mercedes did. They changed the VFC housing design to allow for better air flow around the bms. The redesigned VFC is now 8-mm closer to the engine than before. The bms is mounted on the back of the VFC, the side facing the engine, not the side facing the radiator. So ambient air flow has to reach the bms from the back of the VCF. Much of that ambient air is drawn from the radiator and directed towards the engine block. It is then reflected back to the rear of the VFC where the bms is mounted. By relocating the VFC closer to the engine block, it would then be logical to presume that you will effectively be increasing the air flow to the bms. In this new VFC location, the bms is likely to react more quickly, as now not only is there a greater flow of ambient air, but the heat from the engine block itself should now play a greater role.

Additionally, as Lea points out, Mercedes only changed the VFC design for the models that were affected. The identical VFC which is also used in some other Mercedes models was not changed, as it seems to work properly.

The difference being that, depending on physical engine layouts in various models, in some, the air flow around the bms is adequate, while in others it is restricted.

I am convinced that when Lea installs his new VFC next week, it will work just fine. If you were however to take the same VFC and mount it on Jim's car, it will fail to engage, because of:

a) Different engines,
b) Different physical engine bay layouts,
c) Different air flow patterns,

Phil

Hi Richard

Welcome to the discussion.

Phil

Nice to see you back :) , I thought it was just Jim and I left.... :(


I wish others would comment more often, so please if anyone following this thread wants to comment, then please do join in...it's all rather interesting..... and plenty of new things to learn too...

:cool:

Edit

Still awaiting response from Sandvik/Kanthal. My ISP has informed me that the email has been 'delayed'. First time in 10 years I've ever received that type of message from Postmaster :confused:

This is interesting... whilst sitting here re-reading some of the posts I see that the pages are being re-ordered dynamically. Upon my first visit we had just started page 10, now I see this thread is only running at 8 pages :confused: :confused: :confused:

Page deleting/modifying happening???

For example, take Pberku's comments on page #119:

Pberku's post is number 119 so how can he refers to a post that hasn't happened????? :eek: :eek:

:confused: :confused: :confused:

Lea

Lea, you are right

Yesterday morning (Sunday) we were up to 152 Posts over 11 Pages. This morning (Monday) we are down to 115 Posts and only 8 pages.

For example your current post no 108, was post # 151 yesterday. Jim's post no 142, the one that included an attachment from Mercedes stating that Mercedes modified the design of the VFC, became post no 138 yesterday. Now it totally vanished. So far, from yesterday to today, we lost close to 50 posts, and 3 pages.

You and Jim, had the most number of posts on this topic, and were probably about eaqual. I was closely behind both of you. I now did some quick checking, and discovered that you have close to 50 posts, I have close to 35 posts, and Jim now has only 5 posts left.

Jim is now busy updating Menu #21 ( http://pages.prodigy.net/jforgione/MB_S500.html ) on his own web page. I am sure that as soon as he is finished with his updates, he will return here so that he can shed some light on his "Disappearing threads mystery"

For the record. As of this morning, this is post no 116.

Phil

Well, what's happened to Jim's posts? This is bizarre! We were all engaged in a topic which may bear some new fruit and conclude shortly (with the confirmation of Sandvik and he arrival of my new clutch shortly), so why the need to start deleting posts?

I'm disappointed that this thread will no longer help others in the future as it's now completely disjointed - maybe we should have all used automatic quotes? I've learnt something from this.... always quote the previous post.


I shall continue updating this post as and when necessary and hopefully Jim will return as his comments are as most welcome as anyone's are.

Lea

Lea, I completely agree. I am as frustrated and surprised as you are. I can understand someone editing his previous posts, or maybe even deleting a post that was irrelevant. But for someone to delete 90% of his posts, [close to 50 Posts], is totally unfair to the rest, and very bizarre to say the least.

I truly hope that Jim has a valid explanation.


Phil