Tested the boards with a DC supply.(couldn't wait to hook up the transformers) Both positive and negative work well and are stable. I've adjusted them to 15.00 V excactly and the output stayed constant with input voltages from 17 to 30 V DC and loads from 30 mA to 400 mA.(Less than 10mV deviation)
Next i'll try them with AC and hook up the scope to see if i can find any ripple.
Kees.
Next i'll try them with AC and hook up the scope to see if i can find any ripple.
Kees.
WOW!!! You are quite the generous one, Bob!BobEllis said:
Are you sure you can absorb the $0.07 loss???😀
Dick, Another thought - are D6 and D15 installed correctly?
Dave, I'm just that kind of guy. Don't say I never gave you anything! 🙂
Kees - glad to see Jens and I arent't the only onse who have built this successfully.
Dave, I'm just that kind of guy. Don't say I never gave you anything! 🙂
Kees - glad to see Jens and I arent't the only onse who have built this successfully.
When I get the last missing parts, I will build and test.
Right now I miss the transistors in order to be able to build.
I will use a trimpot and not the fixed voltage devider in the feedback.
\Jens
Right now I miss the transistors in order to be able to build.
I will use a trimpot and not the fixed voltage devider in the feedback.
\Jens
The output voltages I get on the PSU are closer than I reported earlier (my mistake quoting from memory in stead of my notes): 15.18 and 15.13 V. Perfect enough I would say.
Haven't tested yet with a load but I suppose that 'll be all right too.
Greetings,
Dick.
Haven't tested yet with a load but I suppose that 'll be all right too.
Greetings,
Dick.
Bob, since you have plenty of time (insomnia et all),here's another idea for your power control board.
A DC blocking filter for the transformer mains to prevent hum.
(you know,anti parallel diodes and a capacitor).It's cheap and has cured many toriods of hum problems.
Kees.
A DC blocking filter for the transformer mains to prevent hum.
(you know,anti parallel diodes and a capacitor).It's cheap and has cured many toriods of hum problems.
Kees.
Bob,
received the package today(did i mention i live in the swamp?)
Great stuff, customs even found it so interesting that they took a peek. These regulator boards may become highly addictive.
You'll understand flowers are scheduled tomorrow.
My gratitude to you and the mighty Jens, thank you.
received the package today(did i mention i live in the swamp?)
Great stuff, customs even found it so interesting that they took a peek. These regulator boards may become highly addictive.
You'll understand flowers are scheduled tomorrow.
My gratitude to you and the mighty Jens, thank you.
65V has arrived!
I tested 65V version of the positive regulator tonight. Everything seems to work as expected. For this version i used a 9.1V zener (slightly out of its 5% rating at 9.62V) for the reference. Since the feedback resistors were set for a 10V reference I ended up with 62.6V.
I managed to stuff 2 1,000uF 100V caps on the board (18 mm diameter, they don't quite sit straight.) The biggest 100V output caps I could fit were 220 uF. I reduced the current in the differential and limited the reference current to 5 mA. Details on scaling the voltage regulator will be released soon in the form of an assembly guide and spreadsheet. The TO-92 devices were changed to MPSA42/92 to handle the output voltage.
The regulator held 62.6V from no load to 312 mA. However, above 200 mA about 8 mV of noise showed up on the output. My scope is too noisy to discern what it was. I tried running the input voltage up to give the regulator plenty of room to work to no avail. One of my scope test leads gave up the ghost so I did not check the input ripple. The character of the noise did not change with increased input voltage. At or below 150 mA the noise was well within my scope's residual noise.
When below ~150 mA the regulator shows no sign of passing ripple along as long as the DC input voltage is 2.1V higher than the output.
I tested 65V version of the positive regulator tonight. Everything seems to work as expected. For this version i used a 9.1V zener (slightly out of its 5% rating at 9.62V) for the reference. Since the feedback resistors were set for a 10V reference I ended up with 62.6V.
I managed to stuff 2 1,000uF 100V caps on the board (18 mm diameter, they don't quite sit straight.) The biggest 100V output caps I could fit were 220 uF. I reduced the current in the differential and limited the reference current to 5 mA. Details on scaling the voltage regulator will be released soon in the form of an assembly guide and spreadsheet. The TO-92 devices were changed to MPSA42/92 to handle the output voltage.
The regulator held 62.6V from no load to 312 mA. However, above 200 mA about 8 mV of noise showed up on the output. My scope is too noisy to discern what it was. I tried running the input voltage up to give the regulator plenty of room to work to no avail. One of my scope test leads gave up the ghost so I did not check the input ripple. The character of the noise did not change with increased input voltage. At or below 150 mA the noise was well within my scope's residual noise.
When below ~150 mA the regulator shows no sign of passing ripple along as long as the DC input voltage is 2.1V higher than the output.
Hi,
this is real useful.
Some numbers. Assume 2V of ripple gives an average of 1V.
Minimum input voltage =Vout + 0.5*(Vripple) + 2V1
so for 60Vout we need about 63V1 minimum when loaded.
this is real useful.
If we know the ripple on the input then we also know the measured voltage required on the input.
Some numbers. Assume 2V of ripple gives an average of 1V.
Minimum input voltage =Vout + 0.5*(Vripple) + 2V1
so for 60Vout we need about 63V1 minimum when loaded.
A little more testing before work. I cleaned up the test layout and turned off the computer 2 meters away from the test. Funny how that reduced the "residual noise" that I saw on my scope.
At ~200 mA there was a little noise on the output if the AC input was 50V. Increasing the input to 53VAC eliminated it. I left the input at 53VAC and tried ~300 mA (can you tell I am using 100R load resistors?) There was a bit of noise that going up to 58VAC did not eliminate. I only saw ~1V ripple at 300 mA load, not dipping low enough that it should have had an impact.
Holding 58VAC input, increasing the load to 750 mA there was ~10 mV AC on the output.
For a comparison, I hooked up a Kari's board XBOSOZ supply. For those unfamiliar, it is an adaptation of the balanced Zen preamp supply that Nelson Pass published. It uses a CRC input (2200 uF -10R - 2200 uF in my case) a zener string reference and an IRF610 pass device. The output capacitance is more modest than ours, a single 220 uF and a 1 uF film cap. The rectifier I used is the same as what the group buy supply, a 1N4007 bridge with 100 nF across each diode.
As expected with the CRC, it takes more input voltage to operate. At 75 mA it took 55VAC to get clean output. It needed 63.5VAC (the limit of my autoformer/transformer supply) to get clean output when I increased the load to 150 mA. Drawing 200 mA there was about 5 mV ripple. Aside - if you use a Kari supply for your XBOSOZ, I guess this says one per channel although you may be able to get more out of it with more capacitance on both ends.
One interesting thing is that with the Kari board output ripple appeared sinusoidal with some noise superimposed, rather than the triangle dips of the group buy supply.
At ~200 mA there was a little noise on the output if the AC input was 50V. Increasing the input to 53VAC eliminated it. I left the input at 53VAC and tried ~300 mA (can you tell I am using 100R load resistors?) There was a bit of noise that going up to 58VAC did not eliminate. I only saw ~1V ripple at 300 mA load, not dipping low enough that it should have had an impact.
Holding 58VAC input, increasing the load to 750 mA there was ~10 mV AC on the output.
For a comparison, I hooked up a Kari's board XBOSOZ supply. For those unfamiliar, it is an adaptation of the balanced Zen preamp supply that Nelson Pass published. It uses a CRC input (2200 uF -10R - 2200 uF in my case) a zener string reference and an IRF610 pass device. The output capacitance is more modest than ours, a single 220 uF and a 1 uF film cap. The rectifier I used is the same as what the group buy supply, a 1N4007 bridge with 100 nF across each diode.
As expected with the CRC, it takes more input voltage to operate. At 75 mA it took 55VAC to get clean output. It needed 63.5VAC (the limit of my autoformer/transformer supply) to get clean output when I increased the load to 150 mA. Drawing 200 mA there was about 5 mV ripple. Aside - if you use a Kari supply for your XBOSOZ, I guess this says one per channel although you may be able to get more out of it with more capacitance on both ends.
One interesting thing is that with the Kari board output ripple appeared sinusoidal with some noise superimposed, rather than the triangle dips of the group buy supply.
Resistor Values ?
Hi guys, I received my goodies yesterday and plugged in my soldering iron immediately! (Nice packing job Bob, excellent layout Jens, thanks).
I have a couple of questions.
From what I understand, the output voltage is set by the reference diode and resistors R2 and R11 (sch. v2.2)
Refering to schematic v2.2 then, Vout = Vref( (R2+R11)/R11 ) Correct ?
If so, the schematic values work out to 10*(1+3.3)/3.3 = 13.03V ?
Also, there were two 2k resistors in the kit.
If I plug the 2k into the equation for R11, I get 15.0V.
Can someone shed some light for me.
Thanks.
Hi guys, I received my goodies yesterday and plugged in my soldering iron immediately! (Nice packing job Bob, excellent layout Jens, thanks).
I have a couple of questions.
From what I understand, the output voltage is set by the reference diode and resistors R2 and R11 (sch. v2.2)
Refering to schematic v2.2 then, Vout = Vref( (R2+R11)/R11 ) Correct ?
If so, the schematic values work out to 10*(1+3.3)/3.3 = 13.03V ?
Also, there were two 2k resistors in the kit.
If I plug the 2k into the equation for R11, I get 15.0V.
Can someone shed some light for me.
Thanks.
The 2K resistors are, as you suspected, to set the output at 15V.
Notice that the schematic includes a trim pot. There isn't much of a quantity break on them and adding $4 or so to the parts kit cost to get the option of dialing in exactly 15V didn't generate much interest. So I included the 2K to give 15V nominal without a trim pot installed.
If you use the 3K3 and a 10K trim pot you will be able to adjust the voltage upwards from ~14V to within ~2.5V of your raw DC.
Substituting 5K for R11 will give you 12V. If you are shooting for lower operating voltages, you'll want to change the voltage reference and a couple other parts.
I'll post a guide for baseline assembly and scaling the output voltage shortly.
Notice that the schematic includes a trim pot. There isn't much of a quantity break on them and adding $4 or so to the parts kit cost to get the option of dialing in exactly 15V didn't generate much interest. So I included the 2K to give 15V nominal without a trim pot installed.
If you use the 3K3 and a 10K trim pot you will be able to adjust the voltage upwards from ~14V to within ~2.5V of your raw DC.
Substituting 5K for R11 will give you 12V. If you are shooting for lower operating voltages, you'll want to change the voltage reference and a couple other parts.
I'll post a guide for baseline assembly and scaling the output voltage shortly.
A Caution
When using transformer to PSU leads longer than 15 cm some high frequency ringing may occur. This effect is more pronounced at higher load currents. An R-C snubber across the AC input may prevent ringing. We will leave it to you to determine proper values, as they depend on your transformer and lead lengths.
Using the group buy transformer’s 15 cm leads attached directly to the AC terminals and the input voltage adjusted down to 15V with a Variac, no ringing was observed at 600 mA output and very slight ringing was observed at 750 mA. The ringing was approximately 10 mV peak to peak at 900 mA.
The susceptibility to ringing seems to be at its worst just above the minimum input voltage for proper regulator operation. At higher loads I was able to eliminate the ringing by increasing the input voltage to 18VAC. Using a transformer with sufficient current capability and secondary voltage minimizes ringing potential.
At 600 mA or less, 2x15 VAC at 50 VA/12% regulation (the group buy transformers) is sufficient and my PSU was dead quiet. If you must provide more current, a 100 VA transformer, possibly 18VAC should be used.
The high voltage version using MPSA42 and MPSA92 showed no ringing at any input voltage with 70 cm leads from the transformer to the board.
Also, if the load exceeds 500 mA PSU designer predicts the RMS current in the rectifier will exceed the 1N4007's 1A rating. I tested at 900 mA load current for about five minutes without failure.
I am checking to see what higher current diodes will fit. Jens made the holes are larger than standard for this package.
The 1N4007s and group buy transformer will be fine for most users up to four active filter boards per supply (maybe 6 if some features are not used).
Please post your test results for discussion.
When using transformer to PSU leads longer than 15 cm some high frequency ringing may occur. This effect is more pronounced at higher load currents. An R-C snubber across the AC input may prevent ringing. We will leave it to you to determine proper values, as they depend on your transformer and lead lengths.
Using the group buy transformer’s 15 cm leads attached directly to the AC terminals and the input voltage adjusted down to 15V with a Variac, no ringing was observed at 600 mA output and very slight ringing was observed at 750 mA. The ringing was approximately 10 mV peak to peak at 900 mA.
The susceptibility to ringing seems to be at its worst just above the minimum input voltage for proper regulator operation. At higher loads I was able to eliminate the ringing by increasing the input voltage to 18VAC. Using a transformer with sufficient current capability and secondary voltage minimizes ringing potential.
At 600 mA or less, 2x15 VAC at 50 VA/12% regulation (the group buy transformers) is sufficient and my PSU was dead quiet. If you must provide more current, a 100 VA transformer, possibly 18VAC should be used.
The high voltage version using MPSA42 and MPSA92 showed no ringing at any input voltage with 70 cm leads from the transformer to the board.
Also, if the load exceeds 500 mA PSU designer predicts the RMS current in the rectifier will exceed the 1N4007's 1A rating. I tested at 900 mA load current for about five minutes without failure.
I am checking to see what higher current diodes will fit. Jens made the holes are larger than standard for this package.
The 1N4007s and group buy transformer will be fine for most users up to four active filter boards per supply (maybe 6 if some features are not used).
Please post your test results for discussion.
Well, the holes may be on the big side, but they won't accomodate a 1N5404 without drilling. Does anyone have a suggestion for a >1A diode that has leads 1mm dia or smaller?
Just about everyone should have received their order. Is anyone still waiting? The wiki looks like most haven't updated it.
A few more parts kits are avalable, as well as 20 boards. and all the 1500 uf 35V caps anyone might need.
Just about everyone should have received their order. Is anyone still waiting? The wiki looks like most haven't updated it.
A few more parts kits are avalable, as well as 20 boards. and all the 1500 uf 35V caps anyone might need.
I tried to fit 1N5408 schottky's but they need a hole of 1.3 mm.
After drilling there is not enough copper left to properly solder them in.Besides that the body is to big to fit neatly.Perhaps diodes in a TO220 body will fit but they still need the bigger holes.
The advantage of schottky's is that you don't need the snubber caps so that saves a little room.
Bob, did you try to fit ferrite beeds around the transformer leads?
Kees.
After drilling there is not enough copper left to properly solder them in.Besides that the body is to big to fit neatly.Perhaps diodes in a TO220 body will fit but they still need the bigger holes.
The advantage of schottky's is that you don't need the snubber caps so that saves a little room.
Bob, did you try to fit ferrite beeds around the transformer leads?
Kees.
I didn't bother with ferrites once I got good performance with a reasonable lead length.
I'll pick some up and give it a shot if the weather doesn't let me get outside this weekend.
I'll pick some up and give it a shot if the weather doesn't let me get outside this weekend.
- Status
- Not open for further replies.