Rather than tangent the Acoustasonic thread, and with Nigel's question in mind:
"So how many different amps have you seen where the bias is fixed? - it's pretty unusual, and everyone I've seen over the decades had required obtaining the exact correct spares from the manufacturer."
I tried to think of any Fender solid state amps that WERE adjustable bias. SO I scanned through my files. Fender has made MANY solid state models in recent decades. And indeed the use of the TIP142/147 output pair was popular in their design approach, not just their Acoustasonic line.
I could not find any adjustable ones after the mid 1980s. In the TIP142/147 types, we had several models in the Stage series, the Roc Pro models, Ultimate CHorus, Studio 85, SKX100 Princeton 112, The Performer series, Deluxe 112 and 90, BXR100, KXR100, and of course the hugely popular M-80 series, M80 Chorus, M80 Bass, M80 HM. None bear adjustments.
The story is the same in the non-TIP designs as well. And ther are many of those.
Fender lists a part number for the transistors - 28114 and 28115 for TIP142 and TIP147 respectively. (For detailistas, ther are a bunch of extra zeros on the ends of those.)
COnsidering the hundreds of thousands of amps manufactured, Fender must buy the TIP transistors at least in the 100,000 lot, if not million piece orders.
And moving next door, I can;t recall a bias adjustment on a Peavey product other than tube amps. Even the kilowatt stuff.
Marshall 8100 or VS100 Valvestate models lacked adjustment. They used the BDV65 and 64 output devices, which are close enough to the TIP parts that we use the TIPs in repairs there.
Certainly there are many brands and models that DO require adjustment. And my entire discussion involves ONLY MI and Pro Audio. But none of those companies provide part numbers for selected transistors or matching.
If anyone has any interest, Fender has a lot of schematics on their web site support page.
http://www.fender.com/support/articles.php?section=downloads&category=amplifier-schematics
Marshall VS100
http://www.webphix.com/schematic heaven/www.schematicheaven.com/marshallamps/vs100r_100w.pdf
"So how many different amps have you seen where the bias is fixed? - it's pretty unusual, and everyone I've seen over the decades had required obtaining the exact correct spares from the manufacturer."
I tried to think of any Fender solid state amps that WERE adjustable bias. SO I scanned through my files. Fender has made MANY solid state models in recent decades. And indeed the use of the TIP142/147 output pair was popular in their design approach, not just their Acoustasonic line.
I could not find any adjustable ones after the mid 1980s. In the TIP142/147 types, we had several models in the Stage series, the Roc Pro models, Ultimate CHorus, Studio 85, SKX100 Princeton 112, The Performer series, Deluxe 112 and 90, BXR100, KXR100, and of course the hugely popular M-80 series, M80 Chorus, M80 Bass, M80 HM. None bear adjustments.
The story is the same in the non-TIP designs as well. And ther are many of those.
Fender lists a part number for the transistors - 28114 and 28115 for TIP142 and TIP147 respectively. (For detailistas, ther are a bunch of extra zeros on the ends of those.)
COnsidering the hundreds of thousands of amps manufactured, Fender must buy the TIP transistors at least in the 100,000 lot, if not million piece orders.
And moving next door, I can;t recall a bias adjustment on a Peavey product other than tube amps. Even the kilowatt stuff.
Marshall 8100 or VS100 Valvestate models lacked adjustment. They used the BDV65 and 64 output devices, which are close enough to the TIP parts that we use the TIPs in repairs there.
Certainly there are many brands and models that DO require adjustment. And my entire discussion involves ONLY MI and Pro Audio. But none of those companies provide part numbers for selected transistors or matching.
If anyone has any interest, Fender has a lot of schematics on their web site support page.
http://www.fender.com/support/articles.php?section=downloads&category=amplifier-schematics
Marshall VS100
http://www.webphix.com/schematic heaven/www.schematicheaven.com/marshallamps/vs100r_100w.pdf
Taking your standpoint for a moment , that the output devices are not selected .
Then how would Dirkwrights situation have been handled on the production line . The final voltages he posted for the V across the Vbe multipliers for both amps were virtually indentical , indicating that both Vbe multiplier circuits were acting the same .
Would the amp have been kicked back as a rogue and a tech have to change an output device to make it behave .
Then how would Dirkwrights situation have been handled on the production line . The final voltages he posted for the V across the Vbe multipliers for both amps were virtually indentical , indicating that both Vbe multiplier circuits were acting the same .
Would the amp have been kicked back as a rogue and a tech have to change an output device to make it behave .
On the production line, chances are the chassis would have been kicked aside, and a repair person would have thrown a new board in it. Old board onto rummage pile. A couple minutes later, labor wise, the chassis would be back in the queue to go into an amp cab and complete the product.
Boards are VERY cheap to fender. For someone not used to it, discovering their "do not repair list" can be surprising. Fender, like most of their competitors, has a list of lower end products - entry level gear if you prefer - that they would rather replace for a customer than pay us labor for field repairs under warranty. Little practice amps like they throw in with beginner guitars are not a surprise, but some larger amps are on the list. Look up Fender FM212R - could also be referred to as Frontman 212R - it is on the list. If a customer brings me one of those under warranty, I am to verify the problem, and notify Fender, who will send me a new one. SO when entire amps, board, chassis, wooden cabinet, speakers are exchanged as cheaper than repair, imagine what they do in the factory. And that is not a small product.
But I suspect you were asking as to how they would go about fixing it.
Were they to actually repair it in production circumstances, a tech would have troubleshot it and reacted as he saw fit, but wouldn;t spend much time on it. Were it me there, I'd have shorted the bias transistor to see if the outputs behaved - they would have, as dirk discovered.
As I recall, dirk's numbers showed a couple tenths of a volt greater space between opposing driver base voltages. I could be mistaken, but if not, that is more than enough to get the amp hot. I might have been wrong, but I;d have slapped a new bias transistor in it first. If it was an output, I would have probably found the hot one soon enough by checking voltage across each ballast resistor.
But they also would not have me spending a couple hours on it. ON a forum like this, and in basement shops everywhere, a lot of deciding, and fretting over this or that, and uncertainty goes on, and not wanting to order more than we need. IN the world of a place like Fender, no one spends all that cushion time. IN repairs, parts are cheaper than labor BY FAR, and they are all sitting there in stock already. If there is any question as to bias transistor or whacko output device, BAM< just replace all of them. A bias transistor and four output TIPs I can change in a very few minutes. A dollar's worth of parts and no one cares which one was the actual bad part.
I run a pro audio shop, and am authorized Fender, among others. A typical repair shop can spend up to an hour labor before needing special permission to continue. I have a higher rating than most, and can spend up to two hours before asking. Of course, on something like this I would have to explain myself if I claimed that much time. They'd expect me to have this done expeditiously, or know why not. In other words, they'd rather I just shotgun the finals rather than take an extra hour to narrow it down just for the intellectual payoff.
Boards are VERY cheap to fender. For someone not used to it, discovering their "do not repair list" can be surprising. Fender, like most of their competitors, has a list of lower end products - entry level gear if you prefer - that they would rather replace for a customer than pay us labor for field repairs under warranty. Little practice amps like they throw in with beginner guitars are not a surprise, but some larger amps are on the list. Look up Fender FM212R - could also be referred to as Frontman 212R - it is on the list. If a customer brings me one of those under warranty, I am to verify the problem, and notify Fender, who will send me a new one. SO when entire amps, board, chassis, wooden cabinet, speakers are exchanged as cheaper than repair, imagine what they do in the factory. And that is not a small product.
But I suspect you were asking as to how they would go about fixing it.
Were they to actually repair it in production circumstances, a tech would have troubleshot it and reacted as he saw fit, but wouldn;t spend much time on it. Were it me there, I'd have shorted the bias transistor to see if the outputs behaved - they would have, as dirk discovered.
As I recall, dirk's numbers showed a couple tenths of a volt greater space between opposing driver base voltages. I could be mistaken, but if not, that is more than enough to get the amp hot. I might have been wrong, but I;d have slapped a new bias transistor in it first. If it was an output, I would have probably found the hot one soon enough by checking voltage across each ballast resistor.
But they also would not have me spending a couple hours on it. ON a forum like this, and in basement shops everywhere, a lot of deciding, and fretting over this or that, and uncertainty goes on, and not wanting to order more than we need. IN the world of a place like Fender, no one spends all that cushion time. IN repairs, parts are cheaper than labor BY FAR, and they are all sitting there in stock already. If there is any question as to bias transistor or whacko output device, BAM< just replace all of them. A bias transistor and four output TIPs I can change in a very few minutes. A dollar's worth of parts and no one cares which one was the actual bad part.
I run a pro audio shop, and am authorized Fender, among others. A typical repair shop can spend up to an hour labor before needing special permission to continue. I have a higher rating than most, and can spend up to two hours before asking. Of course, on something like this I would have to explain myself if I claimed that much time. They'd expect me to have this done expeditiously, or know why not. In other words, they'd rather I just shotgun the finals rather than take an extra hour to narrow it down just for the intellectual payoff.
Maybe the clue for the Fender is:
Looking at the Marshall VS100 schematic, it looks very lightly biased, probably running close to class C or (A/C), so wouldn't be sensitive to Vbe differences within reason.
I'm getting the idea that a little crossover distortion isn't a big deal with this kind of equipment, so liberties can be taken with the design.
Edit: Oops, cross post with Enzo
So perhaps they are specially selected devices, as someone suggested in the other thread.
Looking at the Marshall VS100 schematic, it looks very lightly biased, probably running close to class C or (A/C), so wouldn't be sensitive to Vbe differences within reason.
I'm getting the idea that a little crossover distortion isn't a big deal with this kind of equipment, so liberties can be taken with the design.
Edit: Oops, cross post with Enzo
Last edited:
Thanks for starting this thread. Since I am going to be a "basement part time for fun" repair shop for the locals, then I'm going to need some help sometimes. The local guy who used to do this hasn't done it in years, and my partner says that many people have broken gear that needs fixing. I'm just doing this for fun and experience, and to be of service to the local live music community. I only charge for parts right now.
Somewhat different to domestic electronics then - for warranty work you get a small fixed payment towards your labour costs - you can't charge per hour, and you can't get special permission to increase the payment.
And things are no doubt different because things are different in different countries, all that on top of industry segment differences. I used to know guys in consumer electronics repair, but not any more.
COmpanies like Fender pay enough for shops like mine. COmpanies like Fender or Peavey or LoudTechnologies (MAckie/Crate/Ampeg) or Yamaha, etc, will pay $50, $60, or more per hour of labor. 20 years ago it wasn;t so rosy. ANy number of companies had flat rate policies, nicer companies at least had major/minor repair rates. As a competent technician, it rarely takes more than an hour to fix something. Exceptions are as expected - intermittants or large things like 32 channel mix boards.
Fender is more interested in getting the customer happy than anything. All these companies do try to do that. So if a tech is up against time limits and not getting anywhere, the Fender guy might ask what the prospects look like. If the guy just needs to do some more work and get it done, they authorize extra time. but if it appears the tech will just be spinning his wheels, then Fender is more likely to send a whole board or subassembly or whatever to get the job done. Or they may chose to simply replace the product.
Godfrey, you may be correct, but I offered that part number because in my mind it just means there are not different part numbers for different grades of part, nor is there any suffix to add for same. And due to the HUGE numbers of these parts they buy, I am just doubtful anyone sits there grading transistors. I could be wrong.
Your remark about crossoverdistortion is the key. Indeed, these are guitar amps and they are not concerned over tiny percents of distortion. It is not hifi. SO bias it cold, so just about any transistor will work in the finals, and who cares if it leaves a bit of a notch at crossover.
dirk, there is no better techer than experience. Good luck with it.
COmpanies like Fender pay enough for shops like mine. COmpanies like Fender or Peavey or LoudTechnologies (MAckie/Crate/Ampeg) or Yamaha, etc, will pay $50, $60, or more per hour of labor. 20 years ago it wasn;t so rosy. ANy number of companies had flat rate policies, nicer companies at least had major/minor repair rates. As a competent technician, it rarely takes more than an hour to fix something. Exceptions are as expected - intermittants or large things like 32 channel mix boards.
Fender is more interested in getting the customer happy than anything. All these companies do try to do that. So if a tech is up against time limits and not getting anywhere, the Fender guy might ask what the prospects look like. If the guy just needs to do some more work and get it done, they authorize extra time. but if it appears the tech will just be spinning his wheels, then Fender is more likely to send a whole board or subassembly or whatever to get the job done. Or they may chose to simply replace the product.
Godfrey, you may be correct, but I offered that part number because in my mind it just means there are not different part numbers for different grades of part, nor is there any suffix to add for same. And due to the HUGE numbers of these parts they buy, I am just doubtful anyone sits there grading transistors. I could be wrong.
Your remark about crossoverdistortion is the key. Indeed, these are guitar amps and they are not concerned over tiny percents of distortion. It is not hifi. SO bias it cold, so just about any transistor will work in the finals, and who cares if it leaves a bit of a notch at crossover.
dirk, there is no better techer than experience. Good luck with it.
Well, thanks to you guys I've learned a lot and I made a friend of mine very happy. That was the second amp I fixed for him. The other was a 300 watt bass head amp. Someone at the factory had put in a 7815 where a 7915 was supposed to be, and one of the main filter caps had bad solder joints. Sixty five cents later the amp was fixed and he was happy.
He'll be receiving the bad gear for me. We will have a form for them to fill out as well. Since I am part time, they have to realize up front that these repairs may take some time.
As I noted on the other thread, one of the quad of outputs was way off spec. It was drawing a lot more current than the others at the same bias point. So no, I don't think these are selected or matched in any way. I don't think power transistors come in hfe grades, like A, B or C, do they?
He'll be receiving the bad gear for me. We will have a form for them to fill out as well. Since I am part time, they have to realize up front that these repairs may take some time.
As I noted on the other thread, one of the quad of outputs was way off spec. It was drawing a lot more current than the others at the same bias point. So no, I don't think these are selected or matched in any way. I don't think power transistors come in hfe grades, like A, B or C, do they?
Last edited:
A few do, I seem to recall back in the days of 2SD424, Toshiba added a little "o" or I forget what the other letter was, for gain range. But that was a very small percentage of parts.
SInce matching criteria were as varied as the circuits the parts were used in, grading was usually done by the OEM using the part. SO when repairing Crown power amps, you will see a number stamped on the transistor cases aside from its type number. You would order replacements from Crown including the grading number, so all the outputs in your amp were of similar spec. But you had to get them from Crown, you couldn;t call up Motorola and order transistors with Crown grading numbers.
Likewise, LoudTech matches MOSFETs for Ampeg power amp stages.
There are some transistors with A,B,C spec differences. The series TIP41/42 and others come to mind.
TIP41 - 40v
TIP41A - 60v
TIP41B - 80v
TIP41C - 100v
Those variants are all in the Motorola transistor guide. It isn;t some after the fact sorting.
SInce matching criteria were as varied as the circuits the parts were used in, grading was usually done by the OEM using the part. SO when repairing Crown power amps, you will see a number stamped on the transistor cases aside from its type number. You would order replacements from Crown including the grading number, so all the outputs in your amp were of similar spec. But you had to get them from Crown, you couldn;t call up Motorola and order transistors with Crown grading numbers.
Likewise, LoudTech matches MOSFETs for Ampeg power amp stages.
There are some transistors with A,B,C spec differences. The series TIP41/42 and others come to mind.
TIP41 - 40v
TIP41A - 60v
TIP41B - 80v
TIP41C - 100v
Those variants are all in the Motorola transistor guide. It isn;t some after the fact sorting.
Often the suffix letter doesn't mean much except a lower price. I worked at a semiconductor company that sold microcontrollers rated at 8MHz, 10MHz and 12MHz. The fact was they all had to pass final test at 12MHz. I needed a 25 amp bridge rectifier late one Saturday. The ones they have at Radio Shack were rated at 50V. I tested the one I bought on my Tek 576 curve tracer. It was good for over 1000V.
From where ever you buy your semiconductors
As you mentioned earlier the A,B, C on TIP series is voltage rating (I've always only stocked the C highest version), but things like BC212L come in numerous gain groupings, if you have Towers International Transistor Selector (and any engineer should) it explains the meaning of the letters and markings.
I seem to have a vague recollections that BC212/182 etc. also used to have paint spots on them signifying groups as well?.
Hi, I am hoping that one of you can help me. I have a Fender Deluxe 112 Plus. The output transistors are getting very hot even without any input and thus the thermostat shuts off the amp. I am suspecting high frequency oscillation but I may be wrong. Another symptom is that when you switch off the amp, it goes off with a "fart" noise. Can someone please point me in right direction?
I'd tend to think not RF, but I too could be wrong. I am guessing you have no scope.
First, is there any DC on the output? Disconnect the speaker, and work without it until the amp is stable. If there is DC, that alone could explain the heat. Taking the load off it should eliminate the current draw that heats them.
Google "RF Probe". You will instantly see a number of very simple circuits to detect RF with a meter. All of them are more or less the same other than details. You can knock together that little circuit to see if there is RF voltage on the amp output.
When I see a solid state amp running hot just sitting there, my first thoughts go to the bias circuit. I suspect both polarities are turning on at the same time. You have four outputs, two and two. I'd expect about 1 volt on the bases of those, two of them negative of course. More than that will cause overheating. Your bias is four diodes - CR5,6,7,8. Do you get something like half a volt across each one?
Are all of your outputs getting hot, or can you tell that it is mainly the two on one end of the row? Not always easy to tell that. Or, look at those four 0.47 ohm 5 watt resistors by the outputs. Measure the voltage across each one. Are they all about the same? And what might that be? If one or two are a lot higher than the others, that means the associated transistor is conducting a lot more curent.
First, is there any DC on the output? Disconnect the speaker, and work without it until the amp is stable. If there is DC, that alone could explain the heat. Taking the load off it should eliminate the current draw that heats them.
Google "RF Probe". You will instantly see a number of very simple circuits to detect RF with a meter. All of them are more or less the same other than details. You can knock together that little circuit to see if there is RF voltage on the amp output.
When I see a solid state amp running hot just sitting there, my first thoughts go to the bias circuit. I suspect both polarities are turning on at the same time. You have four outputs, two and two. I'd expect about 1 volt on the bases of those, two of them negative of course. More than that will cause overheating. Your bias is four diodes - CR5,6,7,8. Do you get something like half a volt across each one?
Are all of your outputs getting hot, or can you tell that it is mainly the two on one end of the row? Not always easy to tell that. Or, look at those four 0.47 ohm 5 watt resistors by the outputs. Measure the voltage across each one. Are they all about the same? And what might that be? If one or two are a lot higher than the others, that means the associated transistor is conducting a lot more curent.
It could be?, high frequency oscillation often causes 'farting noises' when you turn it off.
You really need to stick a scope on the output, this will tell you in seconds if it's oscillating.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.