This amplifier appeared as a project in Radio-Eletcronics magazine, December 1973, complete with PCB design, and boy did that seem thrilling! You can find a PDF of the article online. The schematic in this video is from that magazine article, which is why it's so elegantly drawn, and why there's an annotation next to L1: "SEE TEXT". The article advertises SWTP's complete kit for $150. Don't skimp on the heatsinks, kids!
Two friends and I built Universal Tigers from scratch when I was in high school. We even wound the transformers. We had access to a HP engineering lab. Both of my friends became electrical engineers. I went into software.
The Tiger .01 amp was my first introduction to "Hi-Fi". My brother went through college and a friend introduced him to them. He had a Garrard turntable, PAT-5 Preamp, and the Tiger .01 amps driving a pair of Advent Legacy speakers. The preamp and amps were built from kits. Jethro Tull's "Living in the past" sounded great coming out of that system!
This is fascinating John, as I was listening to you I was also looking at the JAT 501 schematic as well and this helps me learn which is great as my own JAT design is progressing well (I mean my PCB) and I know it's gonna sound awesome ! Your walk throughs are great....cheers.
A few things to watch out for on these amplifiers: R14 is a 10K resistor with 62 volts impressed across it. This means it is dissipating about 400 milli-watts. These resistors are specified as 1/4 watt size so the resistor cooks and tends to increase in value. Best idea is to replace it with a 1 watt resistor. Another problem I remember having on one of mine (assembled from an SWTPC kit) is the grounding lug on the machine screw that holds the power supply bridge rectifier to the chassis. The chassis is heavily anodized and this was preventing the bolt head from making proper contact with the chassis. The cure is to scrape off the anodizing under the bolt head and for good measure I added an external tooth lock washer under the bolt head to make sure the ground lug has a good connection to the chassis. The four 200 ohm 10 watt resistors that are used to voltage divide the signal to the bases of Q12 and Q13 dissipate a lot of heat. They should be spaced above the circuit board by a half inch or so and a bit of fan cooling is a really good idea. Otherwise after long use the circuit board will be damaged. You state that Q14 and Q15 are provide current limiting for the output stage transistors. This is sort of correct but the circuit is more subtle than a simple current limit. It is actually a Safe Operating ARea (SOAR) protection circuit. If you look at the specification sheets for bipolar power transistors you will find a graph for the safe operating area. Basically this graph tells you what combinations of currents and voltages the transistors will withstand and for how long. There can be combinations of current and voltage that are within the transistor's thermal power dissipation limits that will still destroy the transistor (through what is called secondary breakdown) if they exceed the safe operating area. As you mentioned, R39 and R40 provide a current sense signal to the SOAR transistors. You kind of glossed over R43 and R44. Their function is to provide a voltage sense signal for the voltage across the output transistors. When the voltage across the output transistors is high it takes less current through R39 and R40 to trigger the protection circuit so the allowable current limit changes with the voltage across the output transistors. This isn't a big deal if the amplifier is driving a resistive load because in a resistor the current and voltage in in phase. If the current is high through the transistors the voltage is low across them because most of the voltage is dropped across the load. Real world loads (loudspeakers) are not always 100% resistive and in fact can be quite reactive. This is really bad for the output transistors because the phase angle between the current and voltage in the load can result in a high voltage across the output transistors at the SAME TIME they are conducting a high current. (Think power factor - The amplifier is having to generate and dissipate VARS.) A reactive load really fattens out the load line and can cause output transistor failure through secondary breakdown when the safe operating area is exceeded. The protection circuit takes into account BOTH output stage current and voltage and shunts base drive away from the output stage if the safe operating area will be exceeded. I have 4 of these things and once (years ago) when running a loud university residence party I accidentally ran one at high levels into a short circuit for quite some time. It finally tripped the thermal cutout but otherwise suffered no damage due to this protection circuit. Another thing I really didn't like about Dan Meyer's design is the connection to the output transistor base and emitter pins. In my opinion the "slide on" connectors for these pins are a significant weak point and the back to back heat sink design means the pins are in a pretty deep hole so a proper socket cannot be used. The solution is to bolt a 3 lug terminal strip (one lug common to the bolt bracket) under one of the collector terminal bolts and solder short jumper wires to the transistor pins and join them to the circuit board wires at the terminal strip. Yes, this makes changing an output transistor a soldering iron job but I fee better with solder connections Yes, these amplifiers have their quirks but once you get a few things fixed up they can perform reasonably well and quite reliably.
That's an excellent description of the voltage sensing current limiting circuit. I think my Arcam a60 has a similar protection circuit and it always gets hot. However i do wonder if the speakers are just too much for it.
I have a question , what circuit would you recommend for a balanced to unbalanced 2 channel converter? I have a car radio with balanced high level outputs and i want to feed it into a unbalanced only low level car amp. Id like to not use premade solutions nor audio transformers
Don't know if extra matched 1% 4.7k resistors on the vacant legs of the input differential amplifier would help to even up the current flowing. A penalty here is extra noise inherent in the resistors. As you said to make any input stage resistors all 1% metal film. The driver transistors have a RCA typecode. Substitutes I suggest for the driver transistors are the BD 139 - 16/ BD 140 - 16. Either the 2N5551/2N5401 or ZTX 304/ZTX 504 for the input stage transistors. These will have a similar or above voltage rating when compared to the raw supply rail voltage. Seen some discrete component amplifiers on the Circuit Exchange International website. One low power design also includes current mirror stages in its differential input stage. Whether current mirror stages would help here subject to the voltage swing obtained if the input stage is modified in this way.
My friend didn't build it, he got it via eBay a while back. I like the dual differential idea, it fell out of favor for a higher part count and 3dB more noise than a single. These kits were built without matched parts. I really think there is an error yet to be found in this amp, possibly a suttle error in wiring somewhere. I proposed using higher voltage outputs and eliminating half of them and their drivers, but I'm not even going there till we find the existing issue. His other Tigersaurus has never had trouble, it's of a diiferrent vintage by another unknown builder. This amp is very similar to the .01 Tiger design.
I have built an amplifier a few years ago which was in principle according to this symmetric design until Q10 and Q11 as output stage. The transistor matching for the symmetric differential input stage is not such an issue if you use degeneration resistors (R4 to R7).
As ALWAYS, interesting walk through John. My impression is that the amp's complicated nature is due to the high output power and limited capabilities of transistors from the '70s. I admire the designers for working through these issues. I'll stick to an LM1875 until I get more experience to tackle one of these!
Dan Meyer was WAY ahead of his time!! My father built a tiger and I have had and seen and talked about many of them. The Bias stack resistors are a known issue with these.
There was an article about improving this amp , taming the tiger? About 30years ago in audio amateur publication I hope I'm not confusing it with ampzilla.like your review.
Probably the complementary design was needed one voltage gain was needed in the output stage, otherwise it would be difficult to drive as you have to be careful not to have a shoot through current through the base emitter junction
Good morning all - hopefully some people will react to this post/ comment. I've seen a couple of references to this article entitled "A Restoration Odyssey' or "Taming the Flaming Tiger" which was written by a Benjamin Poehland (or similar) and appeared in the old Audio Amateur magazine. I'm genuinely surprised that somebody over in the US hasn't managed to locate the exact issue this was in. My interest in this type of amplifier is that it is the same output format as the current range of SE amplifier boards being manufactured by Avondale Audio here in the UK. If I can find the parts and make the PCB's I fancy building a pair. I would construct it using the heatsink mounting brackets from Naim NAP135 amplifiers - these are machined to take 4 x TO-3 devices and then install it in ModuShop Dissipante type case. I would certainly look to wire it better than the examples I've seen pictures of that's for sure. So come on guys somebody stateside must be able to locate this article to add to the knowledge base Regards Richard
John, if I remember and understood correctly, your JAT501 has no capacitor in parallel with the feedback resistor that links the output to what is in effect the op-amp inverting input. I understand such capacitors are often included to limit high frequency gain and phase shift in such a way that the gain is less than unity at and above the frequency where the output becomes out of phase (the nyquest function?). This prevents RF oscillation. I did not see an explanation in any of your videos on how this capacitor could be omitted, but I would be interested to do so.
He achieves loop stability with a Miller capacitor on the VAS stage. That should be sufficient to give enough phase margin for stability of the main feedback loop.
@@amitghosh6966@ Amit Ghosh The only suggestion is to note how warm the voice coil gets as you increase power with a noncontact IR thermometer but you can't get into the heart of it so the heat will seep out slowly. The more obvious thing is to note how far you can gently move the cone and never exceed that with power. I suppose voice coil resistance would also increase as it warms but you would need to make very sensitive readings to detect this. Also a high power woofer is built to take heat where a lesser one not-so-much, not sure how you would guess at that.
That looks like a class B differential input stage. How is that not going to result in a dead spot near ground? The negative feedback is based on amplification of the distorted signal, so it won't help...
A thought - maybe match the input transistors - very easy to fo these days with one of those cheap component testers ! The resistors may only be 5% ones as well .... .
That's the thing, I was under the same impression, before acquiring a curve tracer. There's more to these than a single result obtained with a tester, especially with NOS ones you need a lot of patience (and transistors) to get a really good match.
@@paulb4661 couldnt you replace with a modern equivalent? even modern transistors from the same batch are going to be waaaay closer to matching then this amp was ever supplied with back in the 70's. And then matching modern ones is just going to be that much easier and closer. Idk just suggesting not a EE .
@@dantheman1998 You are absolutely right, but then you are no longer restoring the piece of kit to its original state, but rather rebuilding it, which often creates its own problems. Differences in pinout and package type, to name but a few. You may even end up with a generator instead of an amp, if you're unlucky. Besides, you wouldn't stick modern, foldable and heated, wing mirrors on a vintage Chevy, even if they did a better job. Some want their toy to serve its purpose, whilst retaining original compliment of ICs.
This amplifier appeared as a project in Radio-Eletcronics magazine, December 1973, complete with PCB design, and boy did that seem thrilling! You can find a PDF of the article online. The schematic in this video is from that magazine article, which is why it's so elegantly drawn, and why there's an annotation next to L1: "SEE TEXT". The article advertises SWTP's complete kit for $150. Don't skimp on the heatsinks, kids!
Two friends and I built Universal Tigers from scratch when I was in high school. We even wound the transformers. We had access to a HP engineering lab. Both of my friends became electrical engineers. I went into software.
The Tiger .01 amp was my first introduction to "Hi-Fi". My brother went through college and a friend introduced him to them. He had a Garrard turntable, PAT-5 Preamp, and the Tiger .01 amps driving a pair of Advent Legacy speakers. The preamp and amps were built from kits.
Jethro Tull's "Living in the past" sounded great coming out of that system!
This is fascinating John, as I was listening to you I was also looking at the JAT 501 schematic as well and this helps me learn which is great as my own JAT design is progressing well (I mean my PCB) and I know it's gonna sound awesome ! Your walk throughs are great....cheers.
Thanks for the rundown John. Hoping to get the amp to you in the coming months for a physical look see. Learned a lot from your video.
A few things to watch out for on these amplifiers:
R14 is a 10K resistor with 62 volts impressed across it. This means it is dissipating about 400 milli-watts. These resistors are specified as 1/4 watt size so the resistor cooks and tends to increase in value. Best idea is to replace it with a 1 watt resistor.
Another problem I remember having on one of mine (assembled from an SWTPC kit) is the grounding lug on the machine screw that holds the power supply bridge rectifier to the chassis. The chassis is heavily anodized and this was preventing the bolt head from making proper contact with the chassis. The cure is to scrape off the anodizing under the bolt head and for good measure I added an external tooth lock washer under the bolt head to make sure the ground lug has a good connection to the chassis.
The four 200 ohm 10 watt resistors that are used to voltage divide the signal to the bases of Q12 and Q13 dissipate a lot of heat. They should be spaced above the circuit board by a half inch or so and a bit of fan cooling is a really good idea. Otherwise after long use the circuit board will be damaged.
You state that Q14 and Q15 are provide current limiting for the output stage transistors. This is sort of correct but the circuit is more subtle than a simple current limit. It is actually a Safe Operating ARea (SOAR) protection circuit. If you look at the specification sheets for bipolar power transistors you will find a graph for the safe operating area. Basically this graph tells you what combinations of currents and voltages the transistors will withstand and for how long. There can be combinations of current and voltage that are within the transistor's thermal power dissipation limits that will still destroy the transistor (through what is called secondary breakdown) if they exceed the safe operating area. As you mentioned, R39 and R40 provide a current sense signal to the SOAR transistors. You kind of glossed over R43 and R44. Their function is to provide a voltage sense signal for the voltage across the output transistors. When the voltage across the output transistors is high it takes less current through R39 and R40 to trigger the protection circuit so the allowable current limit changes with the voltage across the output transistors. This isn't a big deal if the amplifier is driving a resistive load because in a resistor the current and voltage in in phase. If the current is high through the transistors the voltage is low across them because most of the voltage is dropped across the load. Real world loads (loudspeakers) are not always 100% resistive and in fact can be quite reactive. This is really bad for the output transistors because the phase angle between the current and voltage in the load can result in a high voltage across the output transistors at the SAME TIME they are conducting a high current. (Think power factor - The amplifier is having to generate and dissipate VARS.) A reactive load really fattens out the load line and can cause output transistor failure through secondary breakdown when the safe operating area is exceeded. The protection circuit takes into account BOTH output stage current and voltage and shunts base drive away from the output stage if the safe operating area will be exceeded. I have 4 of these things and once (years ago) when running a loud university residence party I accidentally ran one at high levels into a short circuit for quite some time. It finally tripped the thermal cutout but otherwise suffered no damage due to this protection circuit.
Another thing I really didn't like about Dan Meyer's design is the connection to the output transistor base and emitter pins. In my opinion the "slide on" connectors for these pins are a significant weak point and the back to back heat sink design means the pins are in a pretty deep hole so a proper socket cannot be used. The solution is to bolt a 3 lug terminal strip (one lug common to the bolt bracket) under one of the collector terminal bolts and solder short jumper wires to the transistor pins and join them to the circuit board wires at the terminal strip. Yes, this makes changing an output transistor a soldering iron job but I fee better with solder connections
Yes, these amplifiers have their quirks but once you get a few things fixed up they can perform reasonably well and quite reliably.
That's an excellent description of the voltage sensing current limiting circuit. I think my Arcam a60 has a similar protection circuit and it always gets hot. However i do wonder if the speakers are just too much for it.
Regularly an interesting channel
Its so nice seeing you still post
I have a question , what circuit would you recommend for a balanced to unbalanced 2 channel converter? I have a car radio with balanced high level outputs and i want to feed it into a unbalanced only low level car amp. Id like to not use premade solutions nor audio transformers
Interesting video
Thanks for sharing
Always wait for your videos to come out 🤗
Don't know if extra matched 1% 4.7k resistors on the vacant legs of the input differential amplifier would help to even up the current flowing. A penalty here is extra noise inherent in the resistors. As you said to make any input stage resistors all 1% metal film.
The driver transistors have a RCA typecode. Substitutes I suggest for the driver transistors are the BD 139 - 16/ BD 140 - 16.
Either the 2N5551/2N5401 or ZTX 304/ZTX 504 for the input stage transistors. These will have a similar or above voltage rating when compared to the raw supply rail voltage. Seen some discrete component amplifiers on the Circuit Exchange International website. One low power design also includes current mirror stages in its differential input stage. Whether current mirror stages would help here subject to the voltage swing obtained if the input stage is modified in this way.
My friend didn't build it, he got it via eBay a while back. I like the dual differential idea, it fell out of favor for a higher part count and 3dB more noise than a single. These kits were built without matched parts. I really think there is an error yet to be found in this amp, possibly a suttle error in wiring somewhere. I proposed using higher voltage outputs and eliminating half of them and their drivers, but I'm not even going there till we find the existing issue. His other Tigersaurus has never had trouble, it's of a diiferrent vintage by another unknown builder. This amp is very similar to the .01 Tiger design.
brilliant explanations- thx for that. also the hfe or gm ist to match exactly
I have built an amplifier a few years ago which was in principle according to this symmetric design until Q10 and Q11 as output stage. The transistor matching for the symmetric differential input stage is not such an issue if you use degeneration resistors (R4 to R7).
This reminds me of an amplifier circuit found in the circuits appendix of an older RCA transistor manual.
Wow you really know your stuff 😀
As ALWAYS, interesting walk through John. My impression is that the amp's complicated nature is due to the high output power and limited capabilities of transistors from the '70s. I admire the designers for working through these issues. I'll stick to an LM1875 until I get more experience to tackle one of these!
Dan Meyer was WAY ahead of his time!! My father built a tiger and I have had and seen and talked about many of them. The Bias stack resistors are a known issue with these.
I built a number of the Universal Tiger amps. They all worked fine.
There was an article about improving this amp , taming the tiger? About 30years ago in audio amateur publication
I hope I'm not confusing it with ampzilla.like your review.
Probably the complementary design was needed one voltage gain was needed in the output stage, otherwise it would be difficult to drive as you have to be careful not to have a shoot through current through the base emitter junction
Good morning all - hopefully some people will react to this post/ comment. I've seen a couple of references to this article entitled "A Restoration Odyssey' or "Taming the Flaming Tiger" which was written by a Benjamin Poehland (or similar) and appeared in the old Audio Amateur magazine.
I'm genuinely surprised that somebody over in the US hasn't managed to locate the exact issue this was in.
My interest in this type of amplifier is that it is the same output format as the current range of SE amplifier boards being manufactured by Avondale Audio here in the UK. If I can find the parts and make the PCB's I fancy building a pair. I would construct it using the heatsink mounting brackets from Naim NAP135 amplifiers - these are machined to take 4 x TO-3 devices and then install it in ModuShop Dissipante type case.
I would certainly look to wire it better than the examples I've seen pictures of that's for sure.
So come on guys somebody stateside must be able to locate this article to add to the knowledge base
Regards
Richard
Nice
John, if I remember and understood correctly, your JAT501 has no capacitor in parallel with the feedback resistor that links the output to what is in effect the op-amp inverting input. I understand such capacitors are often included to limit high frequency gain and phase shift in such a way that the gain is less than unity at and above the frequency where the output becomes out of phase (the nyquest function?). This prevents RF oscillation. I did not see an explanation in any of your videos on how this capacitor could be omitted, but I would be interested to do so.
He achieves loop stability with a Miller capacitor on the VAS stage. That should be sufficient to give enough phase margin for stability of the main feedback loop.
So where should be the G letter of the schematic connected correctly? I didn't get that part.
Between R28 & R29
@@JohnAudioTech Thank you
Ah bugger it, let's just have Mr Carlson sort this amp out.
Please make video on how to find out any unknown Speaker or Woofer real Wattage
You really can't determine thatby basic electrical measurements, you need to find the Manufacturers specs.
@@hiddenobserver8447 If there's nothing mentioned by manufacturer except the coil resistance then?
@@amitghosh6966@ Amit Ghosh The only suggestion is to note how warm the voice coil gets as you increase power with a noncontact IR thermometer but you can't get into the heart of it so the heat will seep out slowly. The more obvious thing is to note how far you can gently move the cone and never exceed that with power. I suppose voice coil resistance would also increase as it warms but you would need to make very sensitive readings to detect this. Also a high power woofer is built to take heat where a lesser one not-so-much, not sure how you would guess at that.
werent these mono-blocks?
Yes, Only the 215/A was stereo.
ua-cam.com/video/wWfOpNFTmKE/v-deo.html
❤️☺️☺️
That looks like a class B differential input stage. How is that not going to result in a dead spot near ground? The negative feedback is based on amplification of the distorted signal, so it won't help...
Ideally it all runs class A
@@HillsWorkbench The only way it's running class A is if it's biased so that Q1 or Q5 is always on and the other transistor is always off...
@@ameldaquirk3965 You are projecting...
Ah, I thought I "invented" this bootstrapping concept. (I don't mean that literally; I just thought of it but never built it.)
Is rail meant to refer to ground?
No. The positive or negative supplies are called rails.
@@tommost1 Though signal wise, rails are as good as ground.
@@HillsWorkbench Well they are assumed to have zero impedance to ground in small signal analysis... but there is the matter of the voltage on them.
I thought some folk like distortion? Or am I thinking of something else?
A lot of tube lovers like that touch of 2nd harmonic distortion, makes the music seem brighter, sharper. The effect can enhance or degrade music.
A thought - maybe match the input transistors - very easy to fo these days with one of those cheap component testers !
The resistors may only be 5% ones as well .... .
That's the thing, I was under the same impression, before acquiring a curve tracer. There's more to these than a single result obtained with a tester, especially with NOS ones you need a lot of patience (and transistors) to get a really good match.
@@paulb4661 couldnt you replace with a modern equivalent? even modern transistors from the same batch are going to be waaaay closer to matching then this amp was ever supplied with back in the 70's. And then matching modern ones is just going to be that much easier and closer. Idk just suggesting not a EE .
@@dantheman1998 You are absolutely right, but then you are no longer restoring the piece of kit to its original state, but rather rebuilding it, which often creates its own problems. Differences in pinout and package type, to name but a few. You may even end up with a generator instead of an amp, if you're unlucky. Besides, you wouldn't stick modern, foldable and heated, wing mirrors on a vintage Chevy, even if they did a better job. Some want their toy to serve its purpose, whilst retaining original compliment of ICs.
Great video John!