Sorry I'm a little late:
1: try inserting a 33 ohm resistor in front of the emitters of Q5 and Q6.
2: Can you post pictures?
3: Yes, this is class A. The plate will get incredibly hot... Use a fan if you can, this increases the heatsinking ability a lot.
4: Do you have an oscilloscope? If so, try probing the output and posting a trace.
5: If you were able to get a good sine wave working at first, that means we've done something right. Are you sure it's not the power supply? Class A will be drawing more power than A/B.
6: Did you use wirewound bias resistors? I was not able to get an Allison working when using wirewounds; I suspect the coupling between magnetic fields can cause instability in the Allison. (only change them if my other suggestions don't work)
Also:
1: Try the diode protection scheme in the schematic. This will protect the Allison from overdrive.
2: With Klewis' schematic, the outputs are each drawing 2A of current. This is pretty hot.
3: Please post (or at least describe) the current schematic that matches your real-world layout.
4: To protect the Allison when prototyping, bypass D6 and D7. This will limit the power ability of the Allison so you can't possibly overdrive it.
5: In the Allison, the transistors most prone to problems are Q5 and Q6. Please check to see that they are still working.
I'm sorry, this is my fault. I didn't specify any sort of protection for the chip or the Allison. (I don't know how the chip could be damaged though, it has internal overcurrent protection doesn't it? Also, it is an AB chip, while our design is class A) (and delete D1, D2, D3 and D4, they are not needed).
- keantoken
1: try inserting a 33 ohm resistor in front of the emitters of Q5 and Q6.
2: Can you post pictures?
3: Yes, this is class A. The plate will get incredibly hot... Use a fan if you can, this increases the heatsinking ability a lot.
4: Do you have an oscilloscope? If so, try probing the output and posting a trace.
5: If you were able to get a good sine wave working at first, that means we've done something right. Are you sure it's not the power supply? Class A will be drawing more power than A/B.
6: Did you use wirewound bias resistors? I was not able to get an Allison working when using wirewounds; I suspect the coupling between magnetic fields can cause instability in the Allison. (only change them if my other suggestions don't work)
Also:
1: Try the diode protection scheme in the schematic. This will protect the Allison from overdrive.
2: With Klewis' schematic, the outputs are each drawing 2A of current. This is pretty hot.
3: Please post (or at least describe) the current schematic that matches your real-world layout.
4: To protect the Allison when prototyping, bypass D6 and D7. This will limit the power ability of the Allison so you can't possibly overdrive it.
5: In the Allison, the transistors most prone to problems are Q5 and Q6. Please check to see that they are still working.
I'm sorry, this is my fault. I didn't specify any sort of protection for the chip or the Allison. (I don't know how the chip could be damaged though, it has internal overcurrent protection doesn't it? Also, it is an AB chip, while our design is class A) (and delete D1, D2, D3 and D4, they are not needed).
- keantoken
Attachments
Last edited:
When I use schematic from post 266, I get 1.301v when measuring from emmitter to emitter of the output transistors. But, it oscillates...10.204MHz. Interestingly, it sounds quite nice...
When I use schematic from post 270 it makes decent sine and square waves but doesn't generate any bias voltage. I think it's missing the current sources that feed the driver tranistors.
I'll post a working schematic for 270 for comment.
Ken
I thought so too. The Allison may be one of those circuits that isn't affected badly when it oscillates. (Someone also commented in one of my threads that a HF AC signal through a speaker will improve the speaker's operation, though they did not post a source)
If you can't cure this oscillation any other way, try inserting a 1n cap between the bases of Q1 and Q4, then increase or decrease until the optimal value is found (lower is better).
- keantoken
Oscillation rarely affects the sound quality but will destroy the output devices fairly quickly. What is the value of the emitter resistors? It's not posted on the schematic.
Sorry I'm a little late:
I missed your post, must have been typing at the same time. I'll post my schematic and we can discuss from it. Seems schematic from 270 without current source won't work... I see your schematic below. But, if we can focus on 266 first, at least it has current running through itRegarding the chip, it may not be bad, though i changed it out. I was having trouble with one of my probes today and it did the same thing again, (goofy trace) so, I think it was the probe. I also assume all responsiblity for damages
1: try inserting a 33 ohm resistor in front of the emitters of Q5 and Q6.
2: Can you post pictures?
3: Yes, this is class A. The plate will get incredibly hot... Use a fan if you can, this increases the heatsinking ability a lot.
4: Do you have an oscilloscope? If so, try probing the output and posting a trace.
5: If you were able to get a good sine wave working at first, that means we've done something right. Are you sure it's not the power supply? Class A will be drawing more power than A/B.
6: Did you use wirewound bias resistors? I was not able to get an Allison working when using wirewounds; I suspect the coupling between magnetic fields can cause instability in the Allison. (only change them if my other suggestions don't work)
Also:
1: Try the diode protection scheme in the schematic. This will protect the Allison from overdrive.
2: With Klewis' schematic, the outputs are each drawing 2A of current. This is pretty hot.
3: Please post (or at least describe) the current schematic that matches your real-world layout.
4: To protect the Allison when prototyping, bypass D6 and D7. This will limit the power ability of the Allison so you can't possibly overdrive it.
5: In the Allison, the transistors most prone to problems are Q5 and Q6. Please check to see that they are still working.
I'm sorry, this is my fault. I didn't specify any sort of protection for the chip or the Allison. (I don't know how the chip could be damaged though, it has internal overcurrent protection doesn't it? Also, it is an AB chip, while our design is class A) (and delete D1, D2, D3 and D4, they are not needed).
- keantoken
Klewis' earlier schematic shows .32 ohm resistors.
Note: With the Allison most of the oscillation is common-mode or vertical through the outputs, so very little of the total oscillation power actually goes to the speaker (and there are often lots of harmonics, we may even see more 2nd harmonics than fundamental).
I'm glad you're not uncomfortable ginnea-pigging the design, because getting this to work will increase the possibility of others being interested in the Allison. All my suggestions should work with the 266 version (and switching to 270 may not increase sound quality).
- keantoken
??
If so, that's 2 amps per output (1.3V/2=.65/.32=2) and at a rail voltage of 40V, that's 80 watts dissipated at idle per device. They need to be pretty darn good devices on a pretty darn good heatsink to last long.
Am I starting to sound like AndrewT?
??
If so, that's 2 amps per output (1.3V/2=.65/.32=2) and at a rail voltage of 40V, that's 80 watts dissipated at idle per device. They need to be pretty darn good devices on a pretty darn good heatsink to last long.
Am I starting to sound like AndrewT?
Alright, schematic 1 needs the current sources I added to post 283. Without these it won't work (work well, at least...).
MJL: I mean that, if it's the Allison that's oscillating, most of the oscillation is internal. So a small oscillation might be shown at the output, but in reality the oscillation could be up to 500mA pk-pk through the output devices.
- keantoken
MJL: I mean that, if it's the Allison that's oscillating, most of the oscillation is internal. So a small oscillation might be shown at the output, but in reality the oscillation could be up to 500mA pk-pk through the output devices.
- keantoken
??
If so, that's 2 amps per output (1.3V/2=.65/.32=2) and at a rail voltage of 40V, that's 80 watts dissipated at idle per device. They need to be pretty darn good devices on a pretty darn good heatsink to last long.
Am I starting to sound like AndrewT?
darn, I keep doing that...
so the output resistors are .22, supply voltage is 31v
Don't laugh, but, I steam my latte milk to 160deg F and know how hot the pitcher gets, the heat sink was hotter than 160.
keantoken the 49811 provides the current through pins 1 and 3 Jhdr2. This configuration is the one that produces 1.3v across the Re's and occilates at 10Mhz. I'm going to try putting a 1uf cap across the bases of Q1 and Q4(my Q5).
ken
Klewis, how much power output do you need and what is your load? As per MJL, it is advisable to lower your bias resistors.
And yes you are right, I forgot that the 49811 provides bias current for your schematic.
(large caps like 1uF might have too much series inductance, and might actually cause instability. So if it doesn't work, you might try say 100n just in case)
- keantoken
And yes you are right, I forgot that the 49811 provides bias current for your schematic.
(large caps like 1uF might have too much series inductance, and might actually cause instability. So if it doesn't work, you might try say 100n just in case)
- keantoken
regarding caps, I got my units mixed up...
when you say lower the bais resistors, do you mean higher numerically or lower numerically?
Tried .01uf poly cap as shown below. No luck, still oscillates this time at 11Mhz.
Ken
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.