Strongest Gearbox Possible

[Mezzanine]

That is pretty difficult to put a price tag on... Some more info will help: What else will we need to get the transmission built around these wider gears?

I am probably dreaming, but $600-$1000 for the set would definitely get my interest. As I said, difficult to price.



C

[squaab99t]

That is pretty difficult to put a price tag on... Some more info will help: What else will we need to get the transmission built around these wider gears?

I am probably dreaming, but $600-$1000 for the set would definitely get my interest. As I said, difficult to price.



C

2X what Mezzanine mentioned. Since getting a new OEM replacement cluster maybe be more and more difficult. A $1000 for an upgrade maybe in the ballpark. I'd be in.

Dennis

[allessence]

I would have to know how the set is upgraded?
What would be the base line gear set specs, based on pre 89" or post 89" layshaft/cluster, wider gears, etc.

The 89 on layshaft gear with the inner bearing and seperate race is far superior to the earlier design with the bearing running directly on the layshaft.

Problem is the inner race has been really hard to come by, I haven't found a supplier for a new one in 5 years.

So a suitable replacement would have to be found.

[mmoe]

That is pretty difficult to put a price tag on... Some more info will help: What else will we need to get the transmission built around these wider gears?

I am probably dreaming, but $600-$1000 for the set would definitely get my interest. As I said, difficult to price.



C

Yes, you are dreaming. As a point of reference for production gears, you can still get the original Saab gears 1-4 in pre-89 and most of the gears in post '89 brand new (depends on year which one is NLA). That doesn't include a cluster gear and runs around $1000+ for just the four gears minus the synchro hubs and muffs. I'm hopeful to get it down to $2000 including the cluster gear and whatever solution for shifting is settled on, but that's not based on my current conversation with the supplier, just hopeful. We haven't got close to a soft number yet because the actual design of the gearset is open for debate.

I can say that he will make type 8 primaries or $600 per set on a 10 set minimum with 4 straight cut rows of teeth vs.3 helical rows of teeth (they aren't really helical since they are square to the face of the gear, but that's what multiple hobbers I talked with have called it due to the process of cutting them). I'm pretty sure I'll have enough takers on that to get the minimum produced. Personally, I was expecting closer to $1000 per set given the number of difficult operations to produce them and was pleasantly surprised with the pricing. For the record, he could produce them with "helical" offset teeth, but the price would of course reflect the increase in difficulty.

[mmoe]

That is pretty difficult to put a price tag on... Some more info will help: What else will we need to get the transmission built around these wider gears?

I am probably dreaming, but $600-$1000 for the set would definitely get my interest. As I said, difficult to price.



C

2X what Mezzanine mentioned. Since getting a new OEM replacement cluster maybe be more and more difficult. A $1000 for an upgrade maybe in the ballpark. I'd be in.

Dennis

Difficult? I haven't seen one in a long time!

[mmoe]

I would have to know how the set is upgraded?
What would be the base line gear set specs, based on pre 89" or post 89" layshaft/cluster, wider gears, etc.

The 89 on layshaft gear with the inner bearing and seperate race is far superior to the earlier design with the bearing running directly on the layshaft.

Problem is the inner race has been really hard to come by, I haven't found a supplier for a new one in 5 years.

So a suitable replacement would have to be found.

This is a very good question. I have my own preferences, but they may or may not reflect those of the majority here. My personal opinion is that the best gearset by ratio is the earlier set despite their smaller/weaker build. I'm also considering ratios with the idea of type 8 primaries in mind as well, but for higher revving builds and smaller primaries I'd probably still prefer the earlier ratios. The ratios of the earlier cluster gear/1-4 gears combined with an input shaft/constant mesh gear from a '91 and later provides the shortest gearing 1-4 with a larger bump up to 5th.. I also prefer the 9:33 pinion:ring combination with those shorter 1-4 gears to make them a little taller again and have been machining a custom bearing housing to allow for enlarged bearings that closely resemble the '89 and later housing load capacity. The bearings I'm using there are Timken HM88547/10 and M88048/10. The housing will fit into a chillcast case without modification, though it may be steel if the aluminum doesn't test well for thrust against the thinner sidewall to flange area.

I've also been giving a lot of thought to the cluster gear, which in terms of rebuilds provides a dilemma no matter which one you use. The older cluster gears are easier to come by with decent teeth in lower HP N/A models, but the races for the small cluster bearing are usually shot with pits. The bearing also rides on the shaft, which I consider less than ideal as well. The later models are, as you pointed out, difficult to acquire the inner races for and still have pitting on the cluster gear race itself, though not as extensively as the earlier models usually do. My solution for my own transmissions is to use the older gearsets, which I prefer, and I'm now machining the pre '89 cluster gear races out for larger bearings that have their own outer race. The bearing I've chosen also uses an inner race on the 20mm shaft, which prevents wear to the shaft and adds an easily obtainable and replaceable part as does the bearing itself since you replace the race with the needles. The cluster bearing I'm currently testing out is an INA#HK2526 with an inner race INA#IR20x25x26.5. This bearings capacity in terms of PSI rating slightly outspecs the original cluster bearing (by perhaps 10%) while also offering nearly half again the fatigue load limit capacity and I suspect they should perform similarly to or better than the later model cluster bearings with better serviceability (if the cluster gear survives that long).

The bottom line is that I guess that I can see the gears being truly custom and only loosely based on current gearsets. I have some very wild concepts but will do some mockups first to see if they would even fit the case. What design attributes of the currently available gearset would you say are important outside of the obvious needed improvement in strength?

[Jordan]

If you're looking for the seriousness of interest, I think most of us have already laid down a good amount of cash for stuff and finding a handful of people (10?) that will follow through shouldn't be terribly difficult. We already had a successful group buy of 10 or 12 Clutch-type LSDs imported from the UK.

I think the difficulty with rationalizing paying (even a reasonable amount) for a gearset is that the original ones still seam to hold up reasonably well, their limitations are known and they are inexpensive. Certainly not ideal, but still a pretty good value, unless you are Jennifer and have to swap a trans every oil change.

I'm all for progress though and shoring up what would be the weak link in an otherwise pretty bulletproof (CVs/axles next?) drivetrain. If gear spacing was a vast improvement over stock and the reliability was there , I'd definitely be interested down the line.

[Crazyswede]

my .02..... if you are going to have a stout custom gearset made up go the straight cut gear route. Dump the synchros and go with wide straight cut gears in the ratios you want. The gear box will whine like crazy (every gear will sound like reverse) but they will be strong and a hell of a lot easier and cheaper to make then a modified set of synchronized helical gears. Last time I did research on replicating the saab helical gear set the local gear shaper was estimating $1200 for the cutter and another $8000 for the tooling. So that was just about $10,000 before the first set was even in a box.

good luck, interested to see if this goes anywhere

[Luke]

I really doubt that a decent setup could be cobbled together for under $3000, look at motorsports gearsets for makes with way way more market share, VW, Honda, etc and a decent setup is still that much or more.
Its possible that a 4 speed set with spur gears and dog engagement could be put together for around that money, and thats something I'd be interested in as long as the ratios were rally friendly.
I'd be not so interested in a setup with ratios based around a stock ring and pinion and primaries as theres really no way to obtain suitable ratios for rally with 4-5 speeds and the stock 9:35 or 9:33. The options would either be to have a custom type 2 or 3 primary chain setup with say a 30 tooth top cog (smallest that will fit over the bearing) and a 32 or 33 lower cog, or a custom R&P somewhere in the 4.6 to 5.1 range. The R&P is of course the prefered route as it puts less strain on the individual gears that way.

[Mezzanine]

I think Jordan said it pretty well- If you are looking for a group of SAAB enthusiasts who have proven to put their money where their mouth is, this is the place.
I might be dreaming at $1k, but that doesn't mean I am not in at $2k.
As mentioned, the proposal needs some direction. I need to know what kind of gears we are talking about and what it is going to take to build the box. I can tell you that there are going to be two groups interested: road and rally, both with fairly different needs. I would personally like to see a new gearset that let me keep the synchros.
That said, lets see what the consensus of the group is.

[Jordan]

I think you could have a fairly easy shifting syncho-less box.. not sure though. I really think it depends on the # of dogs and the rest of the design. With such a small case to work with you need every mm of gear face (in my totally uneducated opinion) and synchros take up a lot of width...

[mmoe]

If you're looking for the seriousness of interest, I think most of us have already laid down a good amount of cash for stuff and finding a handful of people (10?) that will follow through shouldn't be terribly difficult. We already had a successful group buy of 10 or 12 Clutch-type LSDs imported from the UK.

I think the difficulty with rationalizing paying (even a reasonable amount) for a gearset is that the original ones still seam to hold up reasonably well, their limitations are known and they are inexpensive. Certainly not ideal, but still a pretty good value, unless you are Jennifer and have to swap a trans every oil change.

I'm all for progress though and shoring up what would be the weak link in an otherwise pretty bulletproof (CVs/axles next?) drivetrain. If gear spacing was a vast improvement over stock and the reliability was there , I'd definitely be interested down the line.

I don't have a huge issue with the factory gears either, though the problem is that they are getting very hard to find in decent condition and I say that with the experience of tearing down several a month. The cluster gear in particular sometimes requires dismantling half a dozen transmissions to fine only one worth installing on a full rebuild. If you run a little extra HP and drive hard, but within the legal limits of the law, you probably won't be able to use that same cluster gear when it's time to rebuild and they will be even harder to come by at that point. I'm willing to invest in a solution that provides better longevity and I'm willing to do some legwork to get something put together, but I too would be just as happy to keep building boxes with the older gearset and stockpiling parts for future rebuilds.

[Crazyswede]

I think you could have a fairly easy shifting syncho-less box.. not sure though. I really think it depends on the # of dogs and the rest of the design. With such a small case to work with you need every mm of gear face (in my totally uneducated opinion) and synchros take up a lot of width...

The gearboxes in most offroad motorcycles are straight cut dog boxes. Of course the power, weight, and load ratios are completely different....but its very easy to shift them up or down without using the clutch and without grinding gears. Probably a lot more challenging in a car version. The big drawback to keep in mind with the straight cut gears is going to be the meshing noise. Probably fine for a race car but not something I'd want to listen to if I had it in a road car. I think the Mitsubishi Evo's run a Gemini gearbox that uses straight cut gears and they howl a bit.

[mmoe]

my .02..... if you are going to have a stout custom gearset made up go the straight cut gear route. Dump the synchros and go with wide straight cut gears in the ratios you want. The gear box will whine like crazy (every gear will sound like reverse) but they will be strong and a hell of a lot easier and cheaper to make then a modified set of synchronized helical gears. Last time I did research on replicating the saab helical gear set the local gear shaper was estimating $1200 for the cutter and another $8000 for the tooling. So that was just about $10,000 before the first set was even in a box.

good luck, interested to see if this goes anywhere

Personally, I'd not be willing to spend that kind of money on a solution that doesn't provide all the functionality of the original set. I would not enjoy having spent a couple thousand dollars on gears that give me a headache at highway speed and shift poorly. I'd rather have spent a little more or put more effort into the design end to make them function like a factory box, but with the desired extra strength.

I'll be talking with the shop in the next few days, but I imagine it's going to take a while before a scope of work can be settled on, only after which can a price be generated. Based on seeing how the machines cut the gears (and he has half a dozen very large hobbing machines), I don't think it would cost much more to do helical than straight. He has tons of helical gears he's produced laying around on work tables in the shop, so it's clearly a normal course of work for him.

As for cutters, he may not have any need to purchase the tooling since he may very well already have it. He's got a decent sized wall of shelves full of cutters. Plus, if we're starting from scratch so to speak, I suspect the exact shape of the teeth can be made to work with tooling he already has. It may be far more economical to add a custom reverse gear and idler gear than to try and match the shape of the teeth exactly and pay for specialized cutters if they aren't on his shelf already.

[mmoe]

I really doubt that a decent setup could be cobbled together for under $3000, look at motorsports gearsets for makes with way way more market share, VW, Honda, etc and a decent setup is still that much or more.
Its possible that a 4 speed set with spur gears and dog engagement could be put together for around that money, and thats something I'd be interested in as long as the ratios were rally friendly.
I'd be not so interested in a setup with ratios based around a stock ring and pinion and primaries as theres really no way to obtain suitable ratios for rally with 4-5 speeds and the stock 9:35 or 9:33. The options would either be to have a custom type 2 or 3 primary chain setup with say a 30 tooth top cog (smallest that will fit over the bearing) and a 32 or 33 lower cog, or a custom R&P somewhere in the 4.6 to 5.1 range. The R&P is of course the prefered route as it puts less strain on the individual gears that way.

It's hard to gauge based on product in other makes mostly because in this case I would not be seeking to profit from their production. In your examples, the parts are made, then either sent to a distributor who marks them up followed by a retailer who also marks them up or they go directly from the manufacturer to the retailer who still marks them up. Typically, those markups would nearly double the price at "retail" and sell for somewhere in between. The question is, what are the retailers paying for them at the point of manufacture and that would be the more accurate comparison.

It's also not really an apples to apples comparison in terms of scope. Having worked on a VW 020 transmission out of a Scirocco (mostly the same as all '80s/'90s era VWs), I'd say a Saab gearset is a little less complicated in design and possibly a little more manufacturer friendly particular in looking at how many teeth each gear has to be cut. VWs have maybe half again more teeth than a Saab gearset, so it would be logical to assume it costs more to produce them (the guy I talked with pretty much starts with counting teeth when you hand him a gear). $2000 is probably optimistic, but not entirely out of the realm of possibility depending on how economically they can be designed and how many sets are built, particularly if he already has the tooling on the shelf. That's where I'd like to see the price end up, since as it has been pointed out, the economical nature of just rebuilding the transmissions with existing parts makes it hard to justify spending a whole lot more than that.