Hello Tiger fans: FYI, I have a couple of these listed on eBay...
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=5760726708&ssPageName=STRK:MESE:IT
-Casey Walsh
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=5760726708&ssPageName=STRK:MESE:IT
-Casey Walsh
PB2 said:
Thanks for the reply on identifying my breed of Tigers.
What you have posted makes a lot of sense. I spent a couple of hours yesterday at the Denver Public Library scanning worn out micro fish of Pop Electronics from 70' thru 75' and only came up with the Oct. 1970 construction article on the earlier version. I seem to remember that most of the construction details were in my Audio Amateur issues that I tossed. I had no idea that there were so many versions of the Universal Tiger amp.
Someone needs to put together an on-line SWTPC amplifier memorial museum showing all forms and variations.
Allen
I've since disected the amp and found on the back of the PC board 175-A.
I'm now trying to hunt down obsolete transistors to build two copies. Seems these old RCA and Motorola's are not easy to find!
Thanks again for letting me know what I've got.
Allen
Originally posted by PB2
What you have there is the Universal Tiger MKII (this was also called the 175-A which I think was the number used when ordering the kit) which added two transistors for output stage protection. They decided this was needed and started shipping the new design and discontinued the old design. I think it was some time later when it was written about in the Audio Amateur where a few other points were covered. Upsizing those resistors to 2W is an important upgrade.
I wrote:
"Optimizing the Tiger 175-A" Audio Amateur 3 - 1972
This article covered minor circuit changes to add output protection, and I believe minor bias changes, better diff amp transistors, and increased R12 and R14 to 2W, this was the Universal Tiger MKII version."
I'm now trying to hunt down obsolete transistors to build two copies. Seems these old RCA and Motorola's are not easy to find!
Thanks again for letting me know what I've got.
Allen
Originally posted by PB2
What you have there is the Universal Tiger MKII (this was also called the 175-A which I think was the number used when ordering the kit) which added two transistors for output stage protection. They decided this was needed and started shipping the new design and discontinued the old design. I think it was some time later when it was written about in the Audio Amateur where a few other points were covered. Upsizing those resistors to 2W is an important upgrade.
I wrote:
"Optimizing the Tiger 175-A" Audio Amateur 3 - 1972
This article covered minor circuit changes to add output protection, and I believe minor bias changes, better diff amp transistors, and increased R12 and R14 to 2W, this was the Universal Tiger MKII version."
There's probably next to no point in tracking down those driver transistors with the built in heatsinks... it's a far better bet to just use some nice modern plastic pack transistors with an appropriate heatsink.
I think the consensus was that the original driver transistors and most everything that passed any current was woefully under sized for the application, this includes the old output devices.
_-_- bear
PS. merely increasing the size of the resistors mentioned in the old articles will certainly help, but will not prevent the units from self-frying...
I think the consensus was that the original driver transistors and most everything that passed any current was woefully under sized for the application, this includes the old output devices.
_-_- bear
PS. merely increasing the size of the resistors mentioned in the old articles will certainly help, but will not prevent the units from self-frying...
Yeah, go back in the thread...
The outputs don't do 30 amps AND the requisite voltage!
We're looking at a nominal 100 watt amp with only two devices...the SOA of newer transistors is greater...
When the drivers latch up, it was noted that their current capability was exceeded instantly: *poof*! And all the magic smoke came out.
Either way when it all goes south with this amp the outputs and the drivers all fry instantly... so what do you think?
Have you ever owned these, btw??
_-_-bear
The outputs don't do 30 amps AND the requisite voltage!
We're looking at a nominal 100 watt amp with only two devices...the SOA of newer transistors is greater...
When the drivers latch up, it was noted that their current capability was exceeded instantly: *poof*! And all the magic smoke came out.
Either way when it all goes south with this amp the outputs and the drivers all fry instantly... so what do you think?
Have you ever owned these, btw??
_-_-bear
I realize the potential for melt down is there but I got 20 good years out of the pair. I'm worried that if I radically alter things that I will lose the clean sound they have always had.
Has anyone heard of any problems subbing the NTE180/181 in place of the 4502/802's? Substitutes for the 40409/10's? Even if these are a good fit, I haven't had any luck with the SS1122/23 except for one cross ref list showing ECG397/396 and 2N5416/3439. I read on an earlier post that these may have been a custom run with the "SS" referencing Southwest something.
Allen
Has anyone heard of any problems subbing the NTE180/181 in place of the 4502/802's? Substitutes for the 40409/10's? Even if these are a good fit, I haven't had any luck with the SS1122/23 except for one cross ref list showing ECG397/396 and 2N5416/3439. I read on an earlier post that these may have been a custom run with the "SS" referencing Southwest something.
Allen
Allen, I haven't looked at the circuit in a long time except for a brief peek at the posted early version a week or so ago.
My immediate reaction is that I'd never use general "replacement" devices, unless I had a really good handle on their specs. Which you don't get from the replacement book(s).
The Motorola (or whomever they are these days) 150XX series of output devices are incredibly good and have much larger SOA than the older devices... If I rebuild mine, I'd probably take a good look at that line and pop them in the holes... (I used some 15025s & their complement in a Phase Linear 400 with no problems, and I think the same ones are used in the Crown Macro Tech line also... rugged devices)
You'd probably do well to do the Baker Clamp thing on this amp... I certainly will give that a shot if and when I ever get to them. That and check the Miller compensation cap (I assume it is there) on the VAS stage as I believe was suggested?
If urs worked for 20 years, I wonder what caused them (both?) to fry??
As far as the input transistors, I doubt there is anything terribly special about them... but perhaps someone who is reading this thread will have some other insights into the circuit's requirements besides the hfe & power handling of the devices?
_-_-bear
PS. keep in mind that in that day, transistors with a highish voltage rating were rare, expensive (relatively) and difficult to source - which may account for the special job markings.
My immediate reaction is that I'd never use general "replacement" devices, unless I had a really good handle on their specs. Which you don't get from the replacement book(s).
The Motorola (or whomever they are these days) 150XX series of output devices are incredibly good and have much larger SOA than the older devices... If I rebuild mine, I'd probably take a good look at that line and pop them in the holes... (I used some 15025s & their complement in a Phase Linear 400 with no problems, and I think the same ones are used in the Crown Macro Tech line also... rugged devices)
You'd probably do well to do the Baker Clamp thing on this amp... I certainly will give that a shot if and when I ever get to them. That and check the Miller compensation cap (I assume it is there) on the VAS stage as I believe was suggested?
If urs worked for 20 years, I wonder what caused them (both?) to fry??
As far as the input transistors, I doubt there is anything terribly special about them... but perhaps someone who is reading this thread will have some other insights into the circuit's requirements besides the hfe & power handling of the devices?
_-_-bear
PS. keep in mind that in that day, transistors with a highish voltage rating were rare, expensive (relatively) and difficult to source - which may account for the special job markings.
Thanks for the reply!
Only one has self destructed with a little help from me. I was messing around on the bench one day cleaning it. Had it plugged in with no load and shorted something. As has been well documented, these amps don't need any assistance in blowing up. It's twin got borrowed a couple of years ago by a fellow worker who transferred himself and my amp out of the state. I now want to rebuild the one and construct a second from scratch.
Maybe I'm crazy for wanting to do this but I had a bad experience building a hafler in the early 80's only to have a very harsh grainy sound at any volume. Maybe it wasn't the Hafler's fault but I have'nt wanted to invest that kind of time and effort again with the possibility of a repeat. If you know of any affordable really clean sounding 50-100 watt kits out there that you can recommend, let me know.
Allen
PS- I'll re-read the previous threads for "Baker clamp" and Miller compensation cap on VAS stage. I'll also take your advice on staying away from the generic transistors.
Only one has self destructed with a little help from me. I was messing around on the bench one day cleaning it. Had it plugged in with no load and shorted something. As has been well documented, these amps don't need any assistance in blowing up. It's twin got borrowed a couple of years ago by a fellow worker who transferred himself and my amp out of the state. I now want to rebuild the one and construct a second from scratch.
Maybe I'm crazy for wanting to do this but I had a bad experience building a hafler in the early 80's only to have a very harsh grainy sound at any volume. Maybe it wasn't the Hafler's fault but I have'nt wanted to invest that kind of time and effort again with the possibility of a repeat. If you know of any affordable really clean sounding 50-100 watt kits out there that you can recommend, let me know.
Allen
PS- I'll re-read the previous threads for "Baker clamp" and Miller compensation cap on VAS stage. I'll also take your advice on staying away from the generic transistors.
I'd highly recommend just buying a used pair from ebay or some hamfest/fleamarket... starting from scratch for these is more work than I think they'd be worth... the PCB is essential to build this amp.
As far as from scratch "kits" there are many to choose from. I usually suggest http://www.BorbelyAudio.com as having some of the best circuits and excellent subjective sound.
He has both class A and AB in the power range you are considering.
I think Old Colony has *some* of his published amps as boards or partial kits (board + components, no chassis).
There are a number of others, including a whole range that are being discussed elsewhere on this forum, from basic amp designs imported from Asia through DIYers duplicating the Krell KSA...
Take ur pick!
As far as from scratch "kits" there are many to choose from. I usually suggest http://www.BorbelyAudio.com as having some of the best circuits and excellent subjective sound.
He has both class A and AB in the power range you are considering.
I think Old Colony has *some* of his published amps as boards or partial kits (board + components, no chassis).
There are a number of others, including a whole range that are being discussed elsewhere on this forum, from basic amp designs imported from Asia through DIYers duplicating the Krell KSA...
Take ur pick!
If you were to drive a 90* load at full power the power supply voltage will drop to where the SOA curve starts to fold back.
The MJ802/4502 have 6.67A SOA at 30VDC. With a 50hz~10Khz audio frequency we may increase this to about 10A for duty cycle. If we de-rate 4:1 for temperature and use AGC 2.5A fuses we will be safe. There is also a thermal cut-off.
I drove mine hard into a 4 ohm load and frequently tripped the thermal cut-off.
Mine were newer models that had the current limiting, but it never kicked in (it makes a loud 'pop' driving a speaker). The newer ones also changed the frequency compensation in the driver gain circuit and lowered the bandwidth by a factor of 10, and increased the gain from 2 to 3 (R15, R16 220 ohms). Try a 0.01µF instead of the 1000pF (C3, C4), that helps prevent the overdrive of the outputs.
Use a Baker clamp.
Replace the tantalum input cap (if yours has one) with something else. This made mine oscillate at about 2Mhz.
Tie a 10 ohm resistor in series with the 100 ohm resistor (R11, R13) and run a wire back down to the base of the output transistor from this point.
Common mode conduction (shoot-through) and oscillation seem to be the biggest problems.
The MJ802/4502 have 6.67A SOA at 30VDC. With a 50hz~10Khz audio frequency we may increase this to about 10A for duty cycle. If we de-rate 4:1 for temperature and use AGC 2.5A fuses we will be safe. There is also a thermal cut-off.
I drove mine hard into a 4 ohm load and frequently tripped the thermal cut-off.
Mine were newer models that had the current limiting, but it never kicked in (it makes a loud 'pop' driving a speaker). The newer ones also changed the frequency compensation in the driver gain circuit and lowered the bandwidth by a factor of 10, and increased the gain from 2 to 3 (R15, R16 220 ohms). Try a 0.01µF instead of the 1000pF (C3, C4), that helps prevent the overdrive of the outputs.
Use a Baker clamp.
Replace the tantalum input cap (if yours has one) with something else. This made mine oscillate at about 2Mhz.
Tie a 10 ohm resistor in series with the 100 ohm resistor (R11, R13) and run a wire back down to the base of the output transistor from this point.
Common mode conduction (shoot-through) and oscillation seem to be the biggest problems.
Thanks for the suggestions. Were you working on the 175-A version or the earlier one? I would like to be able to track down suitable substitutes for the transistors that won't cause me to have to alter the rest of the amp around them. Do you think this is possible? If it will take redesigning, I will instead want to try and find a kit (low cost) that is at least as good as the Tiger was.
I, unfortunately do not have anywhere near the expertise you guys have in amplifier theory and design. I can follow assembly and test instructions but that is where it ends.
Thank again
Allen
I, unfortunately do not have anywhere near the expertise you guys have in amplifier theory and design. I can follow assembly and test instructions but that is where it ends.
Thank again
Allen
What a great thread, so I'll drag it back again. I built the 1970ish Universal Tiger from a reprint of the magazine article. This design was the 80W version with no output protection. A friend and I etched the boards from the layout in the article and built two amps. I had also built the commercial kit version for someone in HS or college. My first attempt was fraught with oscillation, to the point where I scrapped the whole chassis and rebuilt it in a huge rack mount chassis that I could easily work on. The amp has served me well for 30 years. It has never failed in music service or on the test bench *unless* I directly shorted the output. I didn't know all that much 30 years ago (some would say little has changed), but I recently did some upgrades to the amp and came to a better understanding of why they can be so troublesome-
1. The original PCB layout uses a dual ground. The length and routing of the grounds (and everything else) is critical. There was tremendous interaction with the output choke, and the exact size most people would wind on a resistor seemed to be the worst case condition. Smaller was better. I cured almost all the stability problems by mounting the PCBs next to each other, and putting a heavy copper star ground right between them, so no ground wire is more than a couple inches long. The wires to the outputs are also very short. Dan Meyer got away with the commercial design because he had several ground points directly to the chassis, a much better HF arrangement, but lousy for low distortion power amps. Use the diagram in Doug Self's amp book, just be sure to do it right near the boards. The usual ground point between the filter caps is a recipe for disaster.
2. The design just before output protection was added has no 1Kohm resistor in series with the base of the input transistor. This was added in the next article and is critical to stability. As mentioned in a post way above, that 220 pF cap directly to the base is bad. Oddly, removing the cap also causes oscillation.
3. The underrated 100 ohms resistors are well known, and need to be 2W parts (I use a series pair of 49.9 ohm large metal films).
4. I run 19000 uF filter caps on each rail, and limit the supply voltage for about 70W, just as described in this thread for reliability.
5. To keep phase shift in the audio band to an absolute minimum, the 220 uF cap in the feedback network needs to be larger. I run 1000 uF, and the amp is flat within +/- 0.1 dB from 20-20K.
6. The original aluminum Motorola outputs may not be all that great- some of mine are (I think- could be something else) prone to rising distortion at high frequencies. Later Motorola parts in steel packages (I know, yecch, but they work) perform better by a factor of three on the THD meter. We also tried some of the similar TI parts, can't remember the number, but I don't think they were as durable- I don't seem to have any left! IMO, the outputs should be reasonably well matched, and that's difficult for the MJ802/4502 because they just tend to have different gains. You could buy hundreds and easily have the bad luck to never get a good match.
7. I tried using a miller cap, rather than the cap across the feedback network. The amp was stable, and measured ok, but sounded awful. I believe the amp needs heavy feedback at high frequencies to control harmonics (I could be full of c**p). I got rid of the miller cap and went back to a mere 50 pF across the feedback, not the 220 pF recommended, and the great sound came back. Dan's later designs included the base resistor on the 1st transistor and something like 60 pF across the feedback, so he was thinking along the same lines.
8. I've had no problem with the thermal feedback via the diodes, but I do pad across the adjustment pot with a resistor so that an intermittent pot won't destroy the amp. (I have a lot of those metal bodied diodes if anybody needs any) Meyer's heat sinking was woefully inadequate. I use a fairly massive U-channel that lets the amp run at full output for a very long time, though probably not forever without a small fan.
IMO, properly debugged, the Universal Tiger can be reliable, sounds great, and isn't terribly complicated to build. Though the later versions had better specs, they seem overly complicated for what they accomplish. If I were building an amp today, I'd keep the output stage, but double up on the devices. That would help distortion and reliability. I'd probably follow Doug Self's thoughts on a current mirrored version of the front end, a follower for the VAS stage, and use a Vbe multiplier. I'd still avoid the miller cap because of the bad taste it left!
edit- BTW, I routinely hot plug and unplug both inputs and outputs with no problem at all.
1. The original PCB layout uses a dual ground. The length and routing of the grounds (and everything else) is critical. There was tremendous interaction with the output choke, and the exact size most people would wind on a resistor seemed to be the worst case condition. Smaller was better. I cured almost all the stability problems by mounting the PCBs next to each other, and putting a heavy copper star ground right between them, so no ground wire is more than a couple inches long. The wires to the outputs are also very short. Dan Meyer got away with the commercial design because he had several ground points directly to the chassis, a much better HF arrangement, but lousy for low distortion power amps. Use the diagram in Doug Self's amp book, just be sure to do it right near the boards. The usual ground point between the filter caps is a recipe for disaster.
2. The design just before output protection was added has no 1Kohm resistor in series with the base of the input transistor. This was added in the next article and is critical to stability. As mentioned in a post way above, that 220 pF cap directly to the base is bad. Oddly, removing the cap also causes oscillation.
3. The underrated 100 ohms resistors are well known, and need to be 2W parts (I use a series pair of 49.9 ohm large metal films).
4. I run 19000 uF filter caps on each rail, and limit the supply voltage for about 70W, just as described in this thread for reliability.
5. To keep phase shift in the audio band to an absolute minimum, the 220 uF cap in the feedback network needs to be larger. I run 1000 uF, and the amp is flat within +/- 0.1 dB from 20-20K.
6. The original aluminum Motorola outputs may not be all that great- some of mine are (I think- could be something else) prone to rising distortion at high frequencies. Later Motorola parts in steel packages (I know, yecch, but they work) perform better by a factor of three on the THD meter. We also tried some of the similar TI parts, can't remember the number, but I don't think they were as durable- I don't seem to have any left! IMO, the outputs should be reasonably well matched, and that's difficult for the MJ802/4502 because they just tend to have different gains. You could buy hundreds and easily have the bad luck to never get a good match.
7. I tried using a miller cap, rather than the cap across the feedback network. The amp was stable, and measured ok, but sounded awful. I believe the amp needs heavy feedback at high frequencies to control harmonics (I could be full of c**p). I got rid of the miller cap and went back to a mere 50 pF across the feedback, not the 220 pF recommended, and the great sound came back. Dan's later designs included the base resistor on the 1st transistor and something like 60 pF across the feedback, so he was thinking along the same lines.
8. I've had no problem with the thermal feedback via the diodes, but I do pad across the adjustment pot with a resistor so that an intermittent pot won't destroy the amp. (I have a lot of those metal bodied diodes if anybody needs any) Meyer's heat sinking was woefully inadequate. I use a fairly massive U-channel that lets the amp run at full output for a very long time, though probably not forever without a small fan.
IMO, properly debugged, the Universal Tiger can be reliable, sounds great, and isn't terribly complicated to build. Though the later versions had better specs, they seem overly complicated for what they accomplish. If I were building an amp today, I'd keep the output stage, but double up on the devices. That would help distortion and reliability. I'd probably follow Doug Self's thoughts on a current mirrored version of the front end, a follower for the VAS stage, and use a Vbe multiplier. I'd still avoid the miller cap because of the bad taste it left!
edit- BTW, I routinely hot plug and unplug both inputs and outputs with no problem at all.
Welcome and thanks for sharing your thoughts. You might have noticed that I simulated the design above and commented that the stability seems OK. I've been thinking since then that it must be something .... layout and wiring, and large signal considerations were my hunch, you've just confirmed it.
It is interesting that many test stability at small and mid signal levels, however we know that these BJT outputs are slow coming out of saturation and therefore large signal with clipping is probably worst case, or at least different. I'm not suggesting you try it with the Tiger since, in my experience, it is likely to blow up.
I did a quick simulation and provided a zip files here, for both the original and just the output with newer devices:
http://www.diyaudio.com/forums/show...&threadid=93637
There is also this thread where we discuss the Plastic Tiger:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=92036
I only just got a single Universal Tiger recently, that does not work, has the wrong output transistors, and it is the version with output protection.
It is interesting that many test stability at small and mid signal levels, however we know that these BJT outputs are slow coming out of saturation and therefore large signal with clipping is probably worst case, or at least different. I'm not suggesting you try it with the Tiger since, in my experience, it is likely to blow up.
I did a quick simulation and provided a zip files here, for both the original and just the output with newer devices:
http://www.diyaudio.com/forums/show...&threadid=93637
There is also this thread where we discuss the Plastic Tiger:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=92036
I only just got a single Universal Tiger recently, that does not work, has the wrong output transistors, and it is the version with output protection.
Universal Tiger Spice Thread Link
Link to spice thread didn't work, l'll try again:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=93637
Pete B.
Link to spice thread didn't work, l'll try again:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=93637
Pete B.
Allen,
many many moons ago, i built the plastic tiger and could not source the SS1122/1123. I got a hold of some people at Motorola and they told me that these were special marked for SWTP. They told me to use MM3007 and MM4007. When you look at the amplifiers that Dan Meyers designed, the MM3005 and MM4005 are showing up. These are a lower breakdown voltage of the same transistors.
Dominique
many many moons ago, i built the plastic tiger and could not source the SS1122/1123. I got a hold of some people at Motorola and they told me that these were special marked for SWTP. They told me to use MM3007 and MM4007. When you look at the amplifiers that Dan Meyers designed, the MM3005 and MM4005 are showing up. These are a lower breakdown voltage of the same transistors.
Dominique
Heres my take on it from another thread:-
Lament for a Tiger
Tiger tiger burning bright,
your output circuit is black as night
Your front end looks great
but its the overall oscillations that we hate
They are, we fear, linked to that Output stage
and that really shows your age
For the culrpit is that common collector
that really screws the transfer vector
We see excess gain and poles galore
and then, just when we thought it safe, even more
no miller capacitor just like Doug told us
this I tell you is why there is a fuss!
So friends, lets ditch this burning design
and go for something a bit more benign.
Lament for a Tiger
Tiger tiger burning bright,
your output circuit is black as night
Your front end looks great
but its the overall oscillations that we hate
They are, we fear, linked to that Output stage
and that really shows your age
For the culrpit is that common collector
that really screws the transfer vector
We see excess gain and poles galore
and then, just when we thought it safe, even more
no miller capacitor just like Doug told us
this I tell you is why there is a fuss!
So friends, lets ditch this burning design
and go for something a bit more benign.
IMO, the problems aren't with the design, but with the physical implementation. This is not a design that rewards long lead lengths and a less-than-perfect understanding of HF and LF grounding. It needs matched devices. Heck, it needs devices they don't even make anymore! Done right, it's reliable and sweet sounding. Done wrong it's a smoking oscillating mess.