Substituting Ge for Si Transistors

classic Classic list List threaded Threaded
48 messages Options
123
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
Thanks Travis, yes I've read through aload of stuff on the threads about Ge and tto be honest It's been almost as confusing as it has been educational  the answers I've had back have helped a lot though, so now I kinda have a vague idea, but only vague one, still not sure about leakage for example, but re-reading stuff does make a little bit more sense each time soit's all good

As for favourites, well I think it's safe to say that we all have our own, but advice from those in the know is always worth taking on board, even if just to start with, can save all sorts of mistakes and problems and I'd rather play it safe to start with and get stuff that works. I can always experiment later if I feel like it.

I'll have a look for the AC125 as suggested, roughly what sort of quantities would you reccomend? and roughly what sort of useful return from is normal for a given quantity? 50% ?
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
In reply to this post by rocket88
Thanks Zach, I appreciate the help from all of you guys, and I hope the studying is going well

I think I've got most of the answers that I was after, and enough info to fill in most of the remaining blanks.........looks like I might be applying for menbership of the dark side at this rate
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Travis
Administrator
In reply to this post by Pavlos
Leakage can be problematic or helpful depending on what you're building. In the fuzz face, low leakage is great because it isn't needed to help bias the transistors, and transistors with lower leakage tend to be much more temperature stable.

However some circuits were designed with leakage in mind, and will need some leakage for the transistors to bias properly. The Maestro FZ-1 and it's derivatives (FZ-1A, MKI ToneBender, Zonk Machine, Selmer Buzztone, etc) were designed to work with leaky transistors, and they won't work properly without leakage. These tend to be the circuits where the Russian germanium isn't as useful, because the Russian transistors are typically low leakage.

But where you get into the real nitty gritty is sortof beyond just hfe and leakage.. That's where the "mojo" comes in lol. In good, honest EXPERIMENTATION, you will find that one transistor type will deliver a more pleasing sound in a given circuit compared to a different transistor type even with identical hfe and leakage readings. Different transistor types vary in frequency response and clipping characteristics which will have a noticable effect on the overall sound.. It doesn't have to be the most expensive transistor, or a certain desirable part number to deliver the best sound to your ears, but it is important to experiment beyond just hfe and leakage numbers..

I don't really want to think about how many thousands of dollars I have spent on transistors, but fuzz is something I'm passionate about and I appreciate the nuances in the sound and I have a lot of fun getting my perfect sound. For me it is a continual process, not something you can attain with a couple bags of transistors. But you have to start somewhere, and it's all good fun
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
Thanks Travis, you've just confirmed what I'd thought I understood about leakage I'll probably have more questions in future but I've learned quite a lot just over this weekend! And I fully agree about experimentation and letting your ears be the final arbiter of what's right. The 'correct' transistors or clipping diodes may well work, and work well, but swapping one or more of them for something else can make a big difference to the sound and feel of a pedal. I found pretty early on (Following advice to use sockets and experiment) that swaping just one transistor for a slightly different spec or model can have a dramatic effect, or just nicely tweak a good sounding pedal into it's 'sweet spot' As you say it's all those other areas of the spec that also come into play and not just the 'headline' ones that are usually mentioned

And to think a year ago I'd never found a fuzz that I got on with, loved plenty of music that they were used  on, but hated the damn things. They were either all paper thin scratchy nails down a blackboard or muffled angry wasps in a dustbin..........and then I discovered this place, breadboarded a few of the really simple circuits just to see how I got on and I've not looked back and thanks to that (Muff fuzz and 1 knob started it all) I've discovered that there are some very......and I mean very good fuzz to be built
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Travis
Administrator
No prob. Personally I prefer to breadboard to choose transistors vs sockets, because I think the final build is more robust when the transistors are soldered to the board. You can solder the transistor into the socket or use glue but that kinda triggers my mild OCD or whatever lol. We've all got our own methods.

Btw to answer your previous question about what QTY you should buy and what % will be usable that's hard to answer. Technically they will all be usable unless they are defective, but some will be just what you're looking for and others you may not find a use for until later on.

From a sealed box of 100 MP20A I may only select half of them (often less) for use in MKII Benders for example but that doesn't mean the rest are useless. The sealed boxes are preferable because you know the seller hasn't already picked what he wanted out of the batch and left you with the scraps. Otherwise you want to buy stuff that has been tested so that you have an idea what you're getting. Beware that the temperature and test equipment used by the seller could produce different readings than what you get at home.

As for the QTY it depends what you're comfortable spending and how much you are planning on building. Obviously you can add more QTY down the line if you find that you really like a certain part, but you will waste more money on shipping by buying multiple smaller orders (especially with international orders like those from the Bulgarian comrade)
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Beaker
In reply to this post by Pavlos
Zach and Travis have made some really good points here. Travis' last post about buying sealed bags/boxes is especially good. Some sellers have already cherry picked the best ones to sell at a premium, so unsealed lots stand a greater chance of being outside the desired gain ranges - especially with AC125 etc. the classic Fuzz Face and Tonebender transistors.

His point about being comfortable with the amount you spend is also very relevant. You could EASILY spend several hundred Dollars/Pounds?Euros buying parts here if you are not careful. If you have that to spare then no problem, but if your budget is limited then I have the following suggestion for you.

Send an email or ebay message to bg-manue and tell him you are a member on here. Ask him to put together a "sample pack" of transistors and diodes and give him a rough budget for it. List the parts you are especially interested in, and that he can include others that he thinks might be suitable.

I'm sure he will come back with a good quantity of top notch parts, at a very favourable price.
He is used to dealing with pedal builders on this and the other forums, and has always gone out of his way to look after us.

This way, you have a reasonable quantity of parts to cover various builds, and as mentioned before, it gives you a chance to find out which parts you like best (for example, Zach really likes the D9D, while I like the D9K).

Then when you run low, you can order what tickles your fancy by the hundred.

Finally, it's not just Ge parts - these guys sell loads of excellent Soviet Silicon parts too, for example the superb KD521 diodes used in the Sovtek Big Muffs!

Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

rocket88
Administrator
just to echo what beaker said about "best" components. it really is up to your ears in the end. the example beaker gave about me loving D9D diodes, and he really likes D9K, it's all in how those diodes distort, and the frequencies they let through. the D9 series of diodes are an amazing group, as each one has a different character and they're relatively cheap so it's a great and easy to test out different ones. if you're going to get some Ge diodes too besides look at the D310, D311, D312, and D18 diodes, as they look awesome (little nuclear bombs) and sound fantastic. the KD521 diodes IMHO is critical to get a russian muff right, and they've become my favorite Si diode over the years.

also again as far as amount to get, you don't have to go full bore and buy 1000 at a time, even i don't get that many of each at a time. you might want to get 100 of 2 kinds and see what you get and if you're happy with them. the only reason i say 100 is that in most cases that's the size of an unopened box. if you get a box of  gt308/it308 or mp20a it's usually like $20-30usd. with the AC125 you have to watch out for the suffix, as it will tell you the gain spread you should get, and they come in packages of 150, usually for $120usd a package, so you may want to wait for those. if you do get an opened package of them get them from orpheus_2005, and i suggest AC125uZ. he's from bulgaria as well and trustworthy, ie they won't be picked through, just a random amount from an unopened package.

can't agree more with sending bg-manue a message on ebay. he's helped us out a lot with sourcing parts, has fantastic prices, and an amazing guy to boot. he's also got another store, i think it's ml-electronics. so you can shop both and he'll take care of you.
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
Travis, I've had those very same thoughts about soldering/ transistors and other parts into their sockets for the sake of robustness, perhaps just the one leg........belt and braces.....just to be certain....

orpheus_2005 was a seller I'd already found actually, looked promising so good to have that confirmed 

I've already messaged bg-manue regarding some of the suggested transistors, and also ml-electronics regarding diodes, he's got boxes of 50/100 (SFD112, D9b, D311) at what look like reasonable prices, and there are so many builds that suggest 2 or 3 Ge diodes at a time that I think it's likely to be far better value for money to go that route and have a healthy stock than buy smaller quantities. Just waiting for replies.

I'm wondering how essential is the AC125 in tonebender? and if so, how many are needed as a minimum and in which positions? They seem to be fairly expensive (Though not insanely so compared to some) and to be mostly available as sets of 2/3 or bulk packs. Can I get away with not using them for now? most realistic I've fond is unsorted lots of 10 from orpheus_2005 for approx £12-14 GBP

Guys, it is as always a pleasure to read your responses, every piece of advice makes sense in it's own way, and your suggestions of what to get and from where are very much appreciated
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Travis
Administrator
Pavlos wrote
Travis, I've had those very same thoughts about soldering/ transistors and other parts into their sockets for the sake of robustness, perhaps just the one leg........belt and braces.....just to be certain....

orpheus_2005 was a seller I'd already found actually, looked promising so good to have that confirmed 

I've already messaged bg-manue regarding some of the suggested transistors, and also ml-electronics regarding diodes, he's got boxes of 50/100 (SFD112, D9b, D311) at what look like reasonable prices, and there are so many builds that suggest 2 or 3 Ge diodes at a time that I think it's likely to be far better value for money to go that route and have a healthy stock than buy smaller quantities. Just waiting for replies.

I'm wondering how essential is the AC125 in tonebender? and if so, how many are needed as a minimum and in which positions? They seem to be fairly expensive (Though not insanely so compared to some) and to be mostly available as sets of 2/3 or bulk packs. Can I get away with not using them for now? most realistic I've fond is unsorted lots of 10 from orpheus_2005 for approx £12-14 GBP

Guys, it is as always a pleasure to read your responses, every piece of advice makes sense in it's own way, and your suggestions of what to get and from where are very much appreciated
Every ToneBender is completely different, so it is important to be specific about which version you're talking about.

The AC125 is not essential for any ToneBender, and was not used in any original ToneBender.

The AC125 has become popular for ToneBenders mainly because it was used by David Main (DAM Stompboxes), however he used the Philips AC125, which is a different brand and country of origin compared to the now ubiquitous Tungsram AC125.

The Tungsram AC125 is available for less money, in larger quantities at this point in time compared to the Philips AC125.

I suggested the Tungsram AC125 as a relatively inexpensive option to have some leakier transistors, which will be essential for Q3 in a ToneBender MKIII (the 3 knob Bender). The AC125 is not essential, but the other transistors we have recommended will not have the leakage needed to bias properly in that specific circuit.
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
hmm, there are rather a few aren't there, I did a list a bit earlier on in the thread and there are (I think?) a couple of MkIII tonebender variants, and a load of MkII/modded MkII as well. So if Q3 is the one where a degree of leakage is needed then with a bit of luck I shouldn't need too many of them? or would all 3 in a 'matched' set be leaky enough to split into 3 different builds?

orpheus has AC125Uz and AC125Kz both Tungsram in unsorted lots of 10, good idea? or bad idea? especially if they are only essential in Q3........ or would these from bg_manu be a better buy http://www.ebay.co.uk/itm/41-pcs-MP41-Russian-Germanium-PNP-Transistor-AC128-AC178-Vintage-NOS-/171769224011

I think I'm good on all the others you've suggested though, looking like MP20/MP20A, MP21A, MP38 from manue, possibly 1T308V from insa_electronics (batch of 20, Looks ok I think?) just got to figure out how many I need and how much I realistically need/want to spend

And I'm glad you mentioned DAM, I'd forgotten about them, a lot of interesting stuff there as well.....oh dear, there isn't a cure is there

And as always, much appreciated
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

rocket88
Administrator
I like MP41A transistors, but you're not going to see enough leakage for a tonebender. I used 2 with hfe around 100 in my oracle and I love it. Travis will corr ft me if I'm wrong, I probably am, but I think leakage of 300uA and up. I quite like sft transistors for Q3, a little on the darker end of the spectrum, and I used one that has leakage of like 600uA in a tonebender variant and it ducking ripped. It's actually the one I made for Kelly of mothership.
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
Thanks Zach, good to know, especially as the prices don't seem too bad for them in lots of 20+ so maybe useful at some point in the future

am I correct that it is only MkIII tonebenders that need Q3 to be leaky? if it is then the AC125 prices maybe don't look quite so scary as long as I don't need to buy loads to get a handful of ones good for Q3

All looking positive though


Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
with regards to leakage for AC125  in Q3 for a MkII tonebender, I found this for tested set of 4, just wondering how suitable they are?

http://www.ebay.co.uk/itm/4x-AC125-TESTED-MATCHED-TUNGSRAM-Ge-Germanium-Transistors-/351892692218?hash=item51ee705cfa:m:mnrVW00xF2FQUYu4PkJ_cJA
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Travis
Administrator
The MKII doesn't need a leaky Q3, it's the the MKIII that does. The MKII would have had leakage on all 3 transistors in the originals, but like the Fuzz Face it isn't really crucial to making the effect work.

I'd pass. Those are pretty much matched, it would be better to have a little spread to experiment with

As you suggested earlier, a bag of 20 should be great for what you're trying to do
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
Thanks Travis, that's just what I thought about Q3, I think I'I might just be starting to get my head round some of this Ge stuff

Not sure if you misinterpreted what I meant about the AC125's being suitable though? I was only planning to use them for Q3 in MkIII's with maybe MP20A / MP21A in Q1 and Q2, so the batch of four would be used in four separate builds. Main thing I was most interested in knowing about them was regarding the amount of leakage, I've seem a few figures mentioned so am not quite sure what is a good or bad upper/lower amount to aim for? Either way I appreciate the advice
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Travis
Administrator
Those are relatively low leakage. The higher hfe may compensate for the lower leakage to make them sound nice.

A typical starting point would be like 200mA and up for that position. No real set in stone rules here, but as I said the 4 nearly matched transistors leaves no real room for experimentation. You're definitely gonna want to play around with different specs in that position
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
no real set in stone rules?........are you sure that germanium isn't just some kind of hoodoovoodoodarkmagik???

Makes sense about the close matched ones being a bit too constant though. But I'm wondering now if anyone has tried a hybrid MkIII with Si in Q3 to try to avoid having to find the right leaky Ge? or is that just a silly idea?




Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Beaker
Why do you think everyone went crazy fo silicon transistors when they became available?

They could not ditch those pesky unreliable and finicky Germaniums fast enough.
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Pavlos
and likewise ditching valves for those nice predictable transistors then wondering why things sounded different, not at all like we're used to, oh yes, that's right, we like some those finicky componets, those less than perfect technologies that have a certain 'something' that just sounds so good......sometimes it's the flaws that deliver the greatest beauty!

All things have their proper place
Reply | Threaded
Open this post in threaded view
|

Re: Substituting Ge for Si Transistors

Travis
Administrator
In reply to this post by Pavlos
People have done a hybrid MKIII where the darlington pair (Q1 and Q2) are replaced with silicon transistors or even a single Si transistor but it's less common to replace Q3 285) silicon. The DAM GB83 Grease Box is probably the only one I know of but the biasing is extensively modified and it isn't really a MKIII anymore
123