Pass DIY Addict
Joined 2000
Paid Member
Excellent! Thanks, Magura. I have a few ceramic fixtures laying around that I can wire together quickly. What kind of behavior should I expect from the bulbs in terms of their brightness over time? Should they burn steadily or will they glow brightly at first and then dim as the caps are done charging?
They may glow a bit while the caps are charging, then they should be down to a dim glow after that.
As it's first time you flick the switch, the caps may need a little time to get up to snuff, so don't be surprised if the bulbs glow for a minute.
Once things has settled, break out your two meters, and hook one to measure the off-set, and one to measure bias, and dial in.
Magura
As it's first time you flick the switch, the caps may need a little time to get up to snuff, so don't be surprised if the bulbs glow for a minute.
Once things has settled, break out your two meters, and hook one to measure the off-set, and one to measure bias, and dial in.
Magura
Pass DIY Addict
Joined 2000
Paid Member
Thanks, Magura! I have two meters that I can use, so that won't be a problem.
The 35,000uF caps are NoS/Surplus stuff that I bough on E-Bay a few weeks ago. I have already spent some time "reforming" them by connecting them to the power supply through a 9k resistor to slowly charge them up for the first time. Each cap charged for about 1 hour. I then measured them a few days later to see which ones held their charges the best and did my best to match their capacity across the rails. I discharged them with a 18v automobile light bulb before connecting the rest of the electronics.
The 35,000uF caps are NoS/Surplus stuff that I bough on E-Bay a few weeks ago. I have already spent some time "reforming" them by connecting them to the power supply through a 9k resistor to slowly charge them up for the first time. Each cap charged for about 1 hour. I then measured them a few days later to see which ones held their charges the best and did my best to match their capacity across the rails. I discharged them with a 18v automobile light bulb before connecting the rest of the electronics.
Pass DIY Addict
Joined 2000
Paid Member
Pass DIY Addict
Joined 2000
Paid Member
It's Alive!
One Aleph-X up and running!
Not able to comment on sound quality right now as I'm running it from an MP3 player with an old crappy test speaker that has $20 worth of 25 year old drivers in it (and they're suffering from foam rot).
I put the 200w worth of light bulbs in series with the primary on the transformer and they glowed brightly at first and then dimmed a bit once the caps were charged. At this point, the rails were only running at 8-9v and I was able to measure about 0.23v across each of the source resistors, so things were looking good. I measure each bank source resistor and verified that bias changed by adjusting the trim pots in V1 and V3. The heat sinks began to get mildly warm after a few minutes.
After this confidence builder, I removed the light bulbs and went directly for 120VAC from the wall outlet. Boy does that transformer buzz! Its a 750va - will exchanging it for a 1000va transformer reduce the amount of buzz?
Power supply rails also dropped from 24.4v unloaded to 18.8v under load with the amp properly biased (source resistors drop 0.5v). The 0R2 resistor in the CRC is running at 51c and drops 1.8v from the rails. I'm really surprised how much voltage it kills - I think I may reduce these to 0R1. I'm getting 20.6v on the input side of the CRC and 18.8v on the output side.
Offset looks pretty good at this point. The amp is biased so I can read a very consistent 0.500v across all source resistors. With the RCA input shorted, I'm getting 0.039v across the speaker terminals. Absolute DC offset runs 0.812 on the pos speaker terminal and 0.775v on the neg speaker terminal.
Resistors R44/45 drop 0.785v and R1/4 drop 0.778v. I cannot measure any voltage drop at all across R2/3 or R42/43.
Heatsinks run 50 at the top (room temp is currently 17c). Measuring the screw/washer on the inside of the chassis that hold the transistor to the heatsink, I get 60c. Measuring the outside of the heatsink between the transistors yields 56c - looks like I have a pretty good thermal bond between the transistors and the sink.
Thanks to Nelson, Grey, William, and everyone else for all that they have shared over the past few years - I've been keeping notes along the way!
Time to clean up and go to bed!
One Aleph-X up and running!
Not able to comment on sound quality right now as I'm running it from an MP3 player with an old crappy test speaker that has $20 worth of 25 year old drivers in it (and they're suffering from foam rot).
I put the 200w worth of light bulbs in series with the primary on the transformer and they glowed brightly at first and then dimmed a bit once the caps were charged. At this point, the rails were only running at 8-9v and I was able to measure about 0.23v across each of the source resistors, so things were looking good. I measure each bank source resistor and verified that bias changed by adjusting the trim pots in V1 and V3. The heat sinks began to get mildly warm after a few minutes.
After this confidence builder, I removed the light bulbs and went directly for 120VAC from the wall outlet. Boy does that transformer buzz! Its a 750va - will exchanging it for a 1000va transformer reduce the amount of buzz?
Power supply rails also dropped from 24.4v unloaded to 18.8v under load with the amp properly biased (source resistors drop 0.5v). The 0R2 resistor in the CRC is running at 51c and drops 1.8v from the rails. I'm really surprised how much voltage it kills - I think I may reduce these to 0R1. I'm getting 20.6v on the input side of the CRC and 18.8v on the output side.
Offset looks pretty good at this point. The amp is biased so I can read a very consistent 0.500v across all source resistors. With the RCA input shorted, I'm getting 0.039v across the speaker terminals. Absolute DC offset runs 0.812 on the pos speaker terminal and 0.775v on the neg speaker terminal.
Resistors R44/45 drop 0.785v and R1/4 drop 0.778v. I cannot measure any voltage drop at all across R2/3 or R42/43.
Heatsinks run 50 at the top (room temp is currently 17c). Measuring the screw/washer on the inside of the chassis that hold the transistor to the heatsink, I get 60c. Measuring the outside of the heatsink between the transistors yields 56c - looks like I have a pretty good thermal bond between the transistors and the sink.
Thanks to Nelson, Grey, William, and everyone else for all that they have shared over the past few years - I've been keeping notes along the way!
Time to clean up and go to bed!
Pass DIY Addict
Joined 2000
Paid Member
Thanks for your help with the turn-on procedure, Magura! It was a bit of a nerve wracking experience to throw the switch for the first time!
I know others have said this, but this amp runs HOT! (Way hotter than my stereo A40 from a few years ago.) After the amp reached a steady state temperature, the buzz from the transformer settled down a bit and once I installed the top and front panels, it was barely audible. Don't know if its worth messing with right now. Given the heat level, I think it may be a good idea to perforate the top panel to let some air through the chassis.
I'm a bit disappointed with the voltage drop in the power supply, but then I've taken specific steps to bring it down by installing two thermistors on the transformer primaries and the 0R2 in the CRC. I'm guessing without them I would probably see about 3v more on the rails which would bring things to ~21-22v like I expected. Given the already high temperature of the chassis, I'm not sure that I really want to raise the rails by 3v as it will make the amp even hotter... I'm undecided whether to leave the thermistors in place or to work on a soft-start circuit... I was really just after a soft start
Ripple currents on the first set of 70,000uF must be fairly high as these caps before the resistor run warm to the touch. The 220,000uF caps after the resistor are not warm.
I cannot bring the absolute DC offset any lower that it is - the trim pot is already adjusted all of the way down (clockwise). If I begin turning it counter clockwise, the absolute DC offset begins to climb right away.
Now its time to copy what I did with the other two amps... That theater room will be heating up more quickly than I thought (its 22 feet x 14 feet x 8 feet).
I know others have said this, but this amp runs HOT! (Way hotter than my stereo A40 from a few years ago.) After the amp reached a steady state temperature, the buzz from the transformer settled down a bit and once I installed the top and front panels, it was barely audible. Don't know if its worth messing with right now. Given the heat level, I think it may be a good idea to perforate the top panel to let some air through the chassis.
I'm a bit disappointed with the voltage drop in the power supply, but then I've taken specific steps to bring it down by installing two thermistors on the transformer primaries and the 0R2 in the CRC. I'm guessing without them I would probably see about 3v more on the rails which would bring things to ~21-22v like I expected. Given the already high temperature of the chassis, I'm not sure that I really want to raise the rails by 3v as it will make the amp even hotter... I'm undecided whether to leave the thermistors in place or to work on a soft-start circuit... I was really just after a soft start
Ripple currents on the first set of 70,000uF must be fairly high as these caps before the resistor run warm to the touch. The 220,000uF caps after the resistor are not warm.
I cannot bring the absolute DC offset any lower that it is - the trim pot is already adjusted all of the way down (clockwise). If I begin turning it counter clockwise, the absolute DC offset begins to climb right away.
Now its time to copy what I did with the other two amps... That theater room will be heating up more quickly than I thought (its 22 feet x 14 feet x 8 feet).
Last edited:
Pass DIY Addict
Joined 2000
Paid Member
The PSU caps are not at all "hot," but warmer than room temp to the touch. I'll measure them them and see how they perform after the amp is up and running tonight.
The bottom plate is heavily perforated with 3/8" (~10mm) diam holes. I will either perforate the lid or use a mesh/screen across the top to help it cool. One good thing about putting the input differential on the heatsink is that it keeps the circuit (bias point/offset/etc) very stable. After just a few minutes of putting the lid on (solid 1/8" thick aluminum sheet) it was just as hot as the heat sinks.
The whole amp draws 3.44A at 121V out of the wall outlet. I was a bit surprised by this - it works out to 416w of power consumption. Perhaps I'm burning too much power off as heat in the power supply with two thermistors and the resistor in the CRC filter...
Eric
The bottom plate is heavily perforated with 3/8" (~10mm) diam holes. I will either perforate the lid or use a mesh/screen across the top to help it cool. One good thing about putting the input differential on the heatsink is that it keeps the circuit (bias point/offset/etc) very stable. After just a few minutes of putting the lid on (solid 1/8" thick aluminum sheet) it was just as hot as the heat sinks.
The whole amp draws 3.44A at 121V out of the wall outlet. I was a bit surprised by this - it works out to 416w of power consumption. Perhaps I'm burning too much power off as heat in the power supply with two thermistors and the resistor in the CRC filter...
Eric
Pass DIY Addict
Joined 2000
Paid Member
I'm trying to determine the actual bias point of the amplifier - can you tell me if I'm going about this the right way?
Using the output stage and ohms law, the source resistors are 0.333 ohm and they are dropping 0.500v, so bias current = 0.500/0.333 = 1.5A per mosfet. With 3 mosfets per bank (each Q in the schematic) this is 4.5A per each of Q1, Q2, Q10, Q11. Thus, I'm drawing 9A per "side" of the balanced amplifier.
I can verify this 9A figure by looking at the CRC filter on each rail: the 0.2 ohm resistor is dropping 1.8v. Using I=V/R I get I=1.8v/0.2 = 9, so each rail is pulling 9A from the PSU.
Also my VOM is telling me I'm pulling 3.44A out of the all socket at 121v - 416watts in total! I guess this verifies the relatively high bias point...
Am I doing this correctly?
Using the output stage and ohms law, the source resistors are 0.333 ohm and they are dropping 0.500v, so bias current = 0.500/0.333 = 1.5A per mosfet. With 3 mosfets per bank (each Q in the schematic) this is 4.5A per each of Q1, Q2, Q10, Q11. Thus, I'm drawing 9A per "side" of the balanced amplifier.
I can verify this 9A figure by looking at the CRC filter on each rail: the 0.2 ohm resistor is dropping 1.8v. Using I=V/R I get I=1.8v/0.2 = 9, so each rail is pulling 9A from the PSU.
Also my VOM is telling me I'm pulling 3.44A out of the all socket at 121v - 416watts in total! I guess this verifies the relatively high bias point...
Am I doing this correctly?
Last edited:
Pass DIY Addict
Joined 2000
Paid Member
Been trying to activate some of the small gray cells.
As I recall, there is no benefit in higher than needed bias an AX.
So you might back off the bias, till you match the needed bias, for supplying the max. current draw, at your specific rail voltage.
Don't forget to take the current gain of the active current source into account.
Magura
As I recall, there is no benefit in higher than needed bias an AX.
So you might back off the bias, till you match the needed bias, for supplying the max. current draw, at your specific rail voltage.
Don't forget to take the current gain of the active current source into account.
Magura
Pass DIY Addict
Joined 2000
Paid Member
Yup, I'm right there with you. Bias was too high and output was being limited by the (lower) rail voltage, so the extra bias was just being burned off as heat (though the power calculation spreadsheet [AXE1-2.xls created by WuffWaff] indicated that output power would peak just below 2 ohms - far too low for my needs...).
I fiddled around again last night. I pulled one of the thermistors from the transformer primaries which gave me back 0.2V on the rails and bypassed the resistors in my CRC filter (now its just C) - this gave me an additional 1.7v back. Then I reduced the bias point - this gave me another 0.5v or so. Together, these changes gave back some voltage in the rails and also reduced the hum from the transformer a bit, though it is still there. The overall result is that now I am operating with 21.5v rails, and a total draw for each side of the amp at ~7.4A. I verified this by measuring the voltage drop across the source resistors and rail voltage. This is the highest bias point that results in increased power output into a 4 ohm load (my target for the speakers I have). Any additional bias just results in extra heat dissipation again. This is the sweet spot for my power supply as output power into 8ohms is voltage limited and power output into 4 ohms is current limited.
As an added benefit of having the amp more reasonably and properly biased is that I have more latitude in adjusting absolute DC offset. Previously, it was close to 1v and that is as low as I could make it - my trim pot wouldn't turn any further. With the bias point reduced, I was able to tweak the absolute DC offset down to 0.023v on the positive speaker terminal and 0.005v on the negative speaker terminal. Relative DC offset across the pos & neg speaker terminals is now very, very stable from cold all the way through fully warmed up at something like 0.012v. I'm glad that I took the time to carefully measure and match all of those transistors!
I was afraid that removing the R from the CRC filter would result in noise at the speaker (potentially as a result of increased power supply ripple), but pressing my ear against the tweeter in my speaker revealed no difference - very, very quiet with or without the R, so I'm not going to worry about it.
Tonight I'll shift my focus to the AC Current Gain setting. I have the information from the Wiki and the relevant threads here. I think I'll target setting it close to 60%. I think the "stock" setting is about 50% and remember Nelson indicating that you can go as high as 66% without and negative effects. The power output spreadsheet indicates that with AC Current Gain set to 60%, I should expect to see 95w into 8ohms and just about 170w into 4ohms before clipping. I'll confirm this later on the test bench at the EE lab on campus where I work.
Thanks again, Magura! I very much appreciate your help and guidance!
Eric
I fiddled around again last night. I pulled one of the thermistors from the transformer primaries which gave me back 0.2V on the rails and bypassed the resistors in my CRC filter (now its just C) - this gave me an additional 1.7v back. Then I reduced the bias point - this gave me another 0.5v or so. Together, these changes gave back some voltage in the rails and also reduced the hum from the transformer a bit, though it is still there. The overall result is that now I am operating with 21.5v rails, and a total draw for each side of the amp at ~7.4A. I verified this by measuring the voltage drop across the source resistors and rail voltage. This is the highest bias point that results in increased power output into a 4 ohm load (my target for the speakers I have). Any additional bias just results in extra heat dissipation again. This is the sweet spot for my power supply as output power into 8ohms is voltage limited and power output into 4 ohms is current limited.
As an added benefit of having the amp more reasonably and properly biased is that I have more latitude in adjusting absolute DC offset. Previously, it was close to 1v and that is as low as I could make it - my trim pot wouldn't turn any further. With the bias point reduced, I was able to tweak the absolute DC offset down to 0.023v on the positive speaker terminal and 0.005v on the negative speaker terminal. Relative DC offset across the pos & neg speaker terminals is now very, very stable from cold all the way through fully warmed up at something like 0.012v. I'm glad that I took the time to carefully measure and match all of those transistors!
I was afraid that removing the R from the CRC filter would result in noise at the speaker (potentially as a result of increased power supply ripple), but pressing my ear against the tweeter in my speaker revealed no difference - very, very quiet with or without the R, so I'm not going to worry about it.
Tonight I'll shift my focus to the AC Current Gain setting. I have the information from the Wiki and the relevant threads here. I think I'll target setting it close to 60%. I think the "stock" setting is about 50% and remember Nelson indicating that you can go as high as 66% without and negative effects. The power output spreadsheet indicates that with AC Current Gain set to 60%, I should expect to see 95w into 8ohms and just about 170w into 4ohms before clipping. I'll confirm this later on the test bench at the EE lab on campus where I work.
Thanks again, Magura! I very much appreciate your help and guidance!
Eric
Last edited:
Pass DIY Addict
Joined 2000
Paid Member