VIN and Engine Number Spread

[kp]

I've started this thread here as I think it is more a general type discussion ---- and it's a long write and..... please note - I am but one who is submitting an idea/thought/ theory/suggestion, and in no way do I mean the following to be an absolute 

For a long while I've been thinking about how some VIN/E combinations are close whilst others are way off. Sure, many pairings are as a result of mixing frames with engines from another set and so on. However there are many pairs that have a large spread yet don't seem to be the norm I will use the earliest such as VIN7 E90, VIN9 E98 and VIN28 E132 as examples (but there are many more) as these 3, I submit, are paired from the factory

The statement that a frame/engine combination can be up to 300 out is really a myth IMHO. I actually don't know where this claim originated from or what evidence there is to support the generally held view. No doubt there are FACTORY frame/engine combinations that this can hold true for BUT, and this is the key, why 300 and not 400 or 500. Nevertheless, there are not that many with wide variation and my belief is that the 300 or higher rule should be used with caution for all large variations. Certainly for the first 300 VIN bikes, such a variation is unlikely. As Dwayne said in another post, the number of recorded VINs in our register is not that many so making any observations is at best, just an observation. Indeed the first VIN we register (not including the aforementioned 3 in para 1) with any wide variation is VIN 179 and of the VIN's recorded up to around 1200 there are only 3 with a variation of over 300. Again not a large sample size.

I have looked right through the registry and as we get higher in the VIN numbering there seems to be some interesting variations. Between VIN 2000 and VIN 2999 there is a good number that the a 300 rule would apply even with the small number we have registered. Are they factory pairings. Assume that 50% are then a 300 rule would be reasonable for some. Some years ago I did an exercise of all the then registered VINs and found that over 50% on E numbers were lower than the VIN number. This is an anomaly we've already discussed I just raise it to refresh members memories.

The fact is that most recorded VINs (again small sample size) have an E pairing very close, with the norm being from 50 to 100 difference where a few are up to 200 with a much lesser number over this variation. Some pairings are well out of wack yet these pairings are said (owner claims state this) are original pairings. If I recall Chris R had one of these odd pairings

So the question is why such variation with some pairings. Hmmmm! That's a real dilemma but there are many such pairs that ARE original factory delivered bikes so why?

There is no recorded answer but I believe there are but 2, maybe 3, possibilities.
1: Pure random selection on the production line. High probability/possibility
2: Faulty Engine on quality control test. High possibility
3. 

I personally think that the #2 is a real possibility and equally real probability given there would have been bikes that failed to pass quality control. Lets be realistic, Honda didn't ship such bikes knowing they had a problem, expecting the dealers to fix

Remember that this was a new engine and new engine assembly line and we do know there were issues such as porosity of the casting. So imagine: a new bike rolls off the production line and during its quality control check an oil leak, engine knock, smoking or any number of issues turns up.

Do they try and repair the engine. I think not
or
Do they take said bike remove the engine and replace it with another. This is the most probable solution but pure conjecture on my part although this currently occurs on modern production lines as I understand

Nevertheless, a fault in an engine on completion from the production line usually results in a replacement being fitted. One can ask why the engine isn't tested prior to installation. My understanding of engine manufacture is that this does not occur for several reasons. All engine assembly conditions/methodologies/trainings are completed during test production phase and the confidence interval for a good engine assembly is high earlier on and becomes very high during later assembly, and testing is better completed during Dyno/Road testing under real load conditions. A problem experienced by all auto manufactures with a new model (certainly in the late 60s 70s) is production teething. Whilst the Japanese had a good track record they were no immune

Anyway I thought this would be a good topic for discussion on a larger membership scale.

[Steve Swan]

KP, thanks for taking time to come up with these excellent observations and theories.  Engines needing rework is certainly a possibility, would explain single digit and low 2 digit frames with high 2 digit and low 3 digit engines.  (that is, over time, as issues in manufacturing processes were identified, one could think it would not be unreasonable to see fewer of these disparate vin pairs.)  We know for a fact the pouring of the pre-7415 cases was a slower process than the pressure die cast cases.  and we know the pre-7415 cases were more prone to internal irregularities as evidenced by grind marks inside these cases, especially in the crankshaft area.

myself, if i understand correctly what you have written, i also don't ascribe to any particular number to the "Fvin/Evin spread" in sandcast production.  as you say, it seems some folks may hold to the "300" theory  what i have noted in the past, is *typically* the Evin is higher than the Fvin -and- as the Fvin increases the Fvin and Evin "spread" increases.  Back in the day, i put quite a bit of time and thought in to studying the vin ranges as people were sending in their vins in a higher volume than these days.  and back in the day, i had maybe a couple hundred vins to study, in contrast to the over 500 we have now.  i remember F4363 having E4540 = 177 difference.  most recently, F1865 having E1973 = 108.  Both these vin sets i am certain are original.  both these vin sets portray what one could describe as "conventional thought" as generally the Evin higher than Fvin and an increasing spread from 108 on a 1xxx vin set to 177 on a 4xxx vin set.    and of course the wildly disparate vin sets such as 7/E90 or 28/E132 don't fit in to any neat picture of a gradually increasing difference between Fvins and Evins.  perhaps in the early days of production, more engines were being assembled than were frames and chassis parts available.......

i guess what could be interesting is with the more or less 500 vin sets we've identified, is to find the set with the widest range and the set with the narrowest range.  i don't know if it is accurate or reliable to "believe" one should expect to see the range gain an increasing spread for every 1025, 50 or 100 vins produced.  and "conventional" thought would lead one to think lower vin ranges should have a narrower spread, and as the vin sets numbers get higher, the spread should widen.  a quick look at 7xxx vins in the vin directory seems to pretty much blow up this theory.  (and there are some vin sets, probably less than more, that are not factory paired.)

another thing that could be evalauted, in the 1st 100 vin sets, look for a pattern in the vin spread -and- identify the outliers.  then in the next 400 vin sets look for a pattern in the vin spread and identify the outliers.  ditto for 500 to 1000.  Ditto every 500 sets up to the end of the LH horn  vins.  then ditto RH horn vins every 500 or 1000 frame sets.  then it would be fun to identify the outliers and see what sort of "range" existed within these outliers as well as identify how many outliers exist.

as you say, we have too small a sample draw any reliable conclusion from.  How many sandcasts actually exist ?  we'll never know.....  i would hazard to guess the total number of sandcasts that actually exist to perhaps be somewhere between double and triple the 500+ we have in the vin directory.  i used to know how many actual vin pairs there were (not counting 'incomplete' sets) but i lost track.

and the other challenge, there has to be, for a number of reasons, some degree of inaccuracy to the vin information we have.

i think the bottom line, when it comes to vin ranges, one cannot arrive at any conclusion to make a hard statement to state as "fact."  as far as i am concerned, the majority of this sandcast stuff is that way.  in the case of identifiable distinctions, certainly i do believe there are distinctions we expect to see in certain vin ranges and distinctions we wouldn't expect to see in various vin ranges.  the fun part about "things sandcast," we are always in for another surprise.  as we know, there have been many !!!! !!!!

[Steve Swan]

i guess what could be interesting is with the more or less 500 vin sets we've identified, is to find the set with the widest range and the set with the narrowest range.  i don't know if it is accurate or reliable to "believe" one should expect to see the range gain an increasing spread for every 1025, 50 or 100 vins produced.  and "conventional" thought would lead one to think lower vin ranges should have a narrower spread, and as the vin sets numbers get higher, the spread should widen.  a quick look at 7xxx vins in the vin directory seems to pretty much blow up this theory.  (and there are some vin sets, probably less than more, that are not factory paired.)

another thing that could be evalauted, in the 1st 100 vin sets, look for a pattern in the vin spread -and- identify the outliers.  then in the next 400 vin sets look for a pattern in the vin spread and identify the outliers.  ditto for 500 to 1000.  Ditto every 500 sets up to the end of the LH horn  vins.  then ditto RH horn vins every 500 or 1000 frame sets.  then it would be fun to identify the outliers and see what sort of "range" existed within these outliers as well as identify how many outliers exist.

Duane seems to have a keen statistical mind, would be interesting to see what he could come up with his study of the vins we've collected.

[Steve Swan]

the other thing i forgot to mention, is lower Evin than Fvin.  again, this could suggest random "picking" or a correction to the cases or......

would be interesting to see how many lower Evin/higher Fvin sets there are in comparison to opposite.

[hondasan]

I'm not going to repeat what has been said above in any detail - the above comments are all valid in my view.

What we do know is that the engine / frame number pairing appears random. As to the range of differences, we could take a stab at it by analysing the 500 or so data sets in the directory to find the likely average difference and maximum. This may well support the "300 rule" as being a reasonable guideline, but of course the data is not always accurate as a consequence of engine changes which have happened, sometimes when bikes have been resurrected from parts.
My suspicion is that factory differences stem from a combination of the day to day randomness of production, and the production rate.

 On day one we could expect the numbers to be close due to presumably the low number of frames and engines available at the point of assembly. This low number of units would result in a low level of random variation and therefore low difference. As time went on and production rate increased, maybe due to the pool of available frames and engines awaiting assembly being larger, the possible difference in numbers due to randomness of pairing would of increased somewhat.

Then we have those larger than typical differences we appear to see? Three early VINS have been mentioned which appear to have much later engines (7/E90, 9/E98, 28/E132). Maybe something caused these three frames to be held back a little out of the typical random build process, by which time the engine numbers typically being fitted had increased somewhat. No idea what the production rate was in the first week, but I could imagine a modest delay (a few days would of been enough) to cause this apparently large difference. As production rates increased, the same few days could of resulted in much larger differences due to the greater throughput of units. For example, I owned 1256 / E665 some years ago. Difference of 591 which always felt to be wrong (replacement motor some time early in life?) The patina of the motor and frame parts, zinc plating, alloy, etc, and 100% correct placement of every single fastener suggested it to be factory.

I guess what I am suggesting is that there was an ongoing randomness of paring, itself affected by production rate, on which would be super-imposed and increasing difference of a delayed unit as production rates increased.

Hope that carries some kind of logic!

What intrigues me is as to whether the first bike finished (pictured in a Swedish magazine I recall) on March 15 1969 was actually VIN 1 fitted with Engine 1. Wonder if they bothered? 

Chris

[Steve Swan]

i'm not able to have an owner's manual in front of me, but IF i recall correctly it seems pictured in the manual was Fvin #1 and Evin #2.......  or maybe this is in the factory workshop manual........

[Sgt.Pinback]

The foto modell in the factory workshop is VIN CB750-1000002 and CB750E-1000004

[4pots1969]

Very interesting discussion but the texts are very long and I have great difficulty with the translator...
All this is related to me a big total overflow problem and "panic" in the production !!
The demand was much stronger than what could produce the assembly lines .... There must be many defects on engines like Kp said...
They had too much pressure and obsession to keep the production they made a lot of defects and errors of numbering...
If you take the time to compare the production of the CB250 and CB350 1968, I am convinced that there is no abnormality of the frame/engine as for Sandcast ...
They had planned a total production of 1,000 bikes to and this is what explains the choice for sand casting and actually overwhelmed by the success they have produced more than 1,200 motorcycles per month.. This is what explains all these mistakes...
It is important not to forget that they have made a huge decision to add engines sandcast in the period Diecast knowing that the Partlist would not be good...
This bike went into production too quickly because Honda had to keep promises not to lose the face its biggest rival KAWASAKI.... For me, It all boils down to this...

[Steve Swan]

The foto modell in the factory workshop is VIN CB750-1000002 and CB750E-1000004

Thanks for clarifying Uli !  my memory ain't what it used to be (at work last night.)

[Steve Swan]

Very interesting discussion but the texts are very long and I have great difficulty with the translator...
All this is related to me a big total overflow problem and "panic" in the production !!
The demand was much stronger than what could produce the assembly lines .... There must be many defects on engines like Kp said...
They had too much pressure and obsession to keep the production they made a lot of defects and errors of numbering...
If you take the time to compare the production of the CB250 and CB350 1968, I am convinced that there is no abnormality of the frame/engine as for Sandcast ...
They had planned a total production of 1,000 bikes to and this is what explains the choice for sand casting and actually overwhelmed by the success they have produced more than 1,200 motorcycles per month.. Who would not have made a mistake???It is important not to forget that they have made a huge decision to add engine sandcast in the period Diecast knowing that the Partlist would not be good...
It is important not to forget that they have made a huge decision to add engines sandcast in the period Diecast knowing that the Partlist would not be good...
This bike went into production too quickly because Honda had to keep promises not to lose the face its biggest rival KAWASAKI.... For me, It all boils down to this...

Gerard, you make some great points.  Where did you learn "They had planned a total production of 1,000 bikes..." ?

i have not looked at this website in years, i found it again. very interesting, doubtful the mockup exists - http://www.motorcycleclassics.com/classic-japanese-motorcycles/kawasaki-z1-prototype.aspx

[CBman]

steve: " i have not looked at this website in years, i found it again. very interesting, doubtful the mockup exists - http://www.motorcycleclassics.com/classic-japanese-motorcycles/kawasaki-z1-prototype.aspx "

Yes, if Kawasaki was little quicker, maybe the icon will be that Z750. Not sure if that is true, but I have read in some book that Kawasaki wanted to introduce the bike in beginning of 1969, but Honda was quicker - late 1968. So Kawasaki decided to stop this project and to develop and introduced something better - Z1 900........

Several months ago Kawasaki introduced Ninja H2R, which will be very difficult to overtake, even for Honda.

[Steve Swan]

not to hijack this thread, i guess i'm a honda guy....  i remember seeing the new Z1's in 1973 on the showroom of JerryCo Motors in Lincoln, Nebraska.  i remember thinking thet did not look even nearly as nice as the CB750, but by that time i had "gone british," so the 73 Tridents looked far better than the Z1 or the CB750.

[kp]

Steve and Chris. Great comments on this topic. I'm thinking it would be great to talk to somebody who was actually on the production team at the time I doubt I'll see that happen although I will ask a Japanese friend if he can shed any light on this subject. It is very difficult to have a long conversation due to the language barrier. As Gerard said, long discussions and difficulty in translation although you did OK Gerard with the translation.
The story of Kawasaki is that they were not in a position to get the win over Honda as their project was well behind that of the CB750. Kawasaki were initially doing a 750 as this was the limit for motorcycle size in Japan but whith the CB750 release KHI made the decision to produce the 900 to try to gain an advantage over Honda. I'm not sure of the date that mock up was released but I think 69 is closer to the mark than 68 which indicates KHI were in early development at that time. I always thought rocket 3 pipe design they used was a bit much

[Steve Swan]

"I always thought rocket 3 pipe design they used was a bit much."

looked like shit then.  pure sex now.

[Marcello Tha]

I have F6457 and E7083, 626 range. I am 100% certain that this is the original motor. 1) The motorcycle arrived in Brazil  0km in early 1970 . 2) There is only this Sandcast in Brazil.