Pricey ........ worth it ...?
That remains to be heard! He sent me the two pcbs with a few nice parts, like paper oil caps for $50, and he got me two toroidal transformers made with russian military grade materail (220V: 39V, 58V) for £150 including shipping. And he gave me many hints.
Last edited:
The layout of that board looks a lot like the one created by member PWatts for the KSA-100MKII clone group buy. The placement of R36 and R37 were discussed and could be improved, but that board has them in their original locations. The seller did not make any effort to create his own version.
I have my doubts about the so called superioty of the other boards compaired to the ones from Jims Audio. If that would be the case, the original KSA-100 is inferior too, because Jims boards are exactely the same... My personal consideration and decision to go for Jims boards was that they are just like the original ones, and that was what I wanted, an orinal copy . And to be fair, they are great quality, thick tracks and gold plated contacts, they look more rigide than the other ones to me.
. And to be fair, they are great quality, thick tracks and gold plated contacts, they look more rigide than the other ones to me.Jim's Audio 50/100 MK II BOARDS
Here is a completed KSA 50. See for yourself. http://www.diyaudio.com/forums/solid-state/62168-my-completed-krell-ksa-50-clone-pics-2.htmlPricey ........ worth it ...?
Unfortunatelly no further progress made, so no pictures, but I have a few questions about the build instead First, Jims Audio boards use 2SJ313 / 2SK1023 instead of VP/VN0210N5. When I compare the datasheets, I see little difference, but to be fair I dont really know which parameter is more important than the other. Are there advantages or disadvantages to use one or another? Is the 2SJ313 / 2SK1023 preferable above Zetex ZVP3310 and ZVN3310 which are used in other builds, and why?
I also wanted to ask if it is sane to use fuses in the secondairy rails. I’ve seen designs in which secondairy fuses are used. Mostly after the bridge rectifier and the electrolytic capacitors, so charge pulses caused by charging the capacitors would not destroy the fuses. When you are already using a primairy fuse, is it neccesairy to use them in the secondairies also?
My last question is about the zener diodes. I’ve red the following in the 'big' KSA-100 threat: ‘Keep the zeners to 400mW or 500mW singles. Duals will need twice as much current to get around the knee of the IV curve. 1.3W types will need three times the current (have pity on the dropping resistor and internal temperatures).’ I can understand that when you set Zeners in series they will eat more current. But is there a difference in Zener current when you use a 1.3 W 39V zener or a 400mW zener?
I've uploaded the datasheets of the FET's here:
http://michaelq.home.xs4all.nl/Datasheets/2SJ313.pdf
http://michaelq.home.xs4all.nl/Datasheets/2SK2013.pdf
http://michaelq.home.xs4all.nl/Datasheets/VN0210N5.pdf
http://michaelq.home.xs4all.nl/Datasheets/VP0210N5.pdf
First, Jims Audio boards use 2SJ313 / 2SK1023 instead of VP/VN0210N5. When I compare the datasheets, I see little difference, but to be fair I dont really know which parameter is more important than the other. Are there advantages or disadvantages to use one or another? Is the 2SJ313 / 2SK1023 preferable above Zetex ZVP3310 and ZVN3310 which are used in other builds, and why?I also wanted to ask if it is sane to use fuses in the secondairy rails. I’ve seen designs in which secondairy fuses are used. Mostly after the bridge rectifier and the electrolytic capacitors, so charge pulses caused by charging the capacitors would not destroy the fuses. When you are already using a primairy fuse, is it neccesairy to use them in the secondairies also?
My last question is about the zener diodes. I’ve red the following in the 'big' KSA-100 threat: ‘Keep the zeners to 400mW or 500mW singles. Duals will need twice as much current to get around the knee of the IV curve. 1.3W types will need three times the current (have pity on the dropping resistor and internal temperatures).’ I can understand that when you set Zeners in series they will eat more current. But is there a difference in Zener current when you use a 1.3 W 39V zener or a 400mW zener?
I've uploaded the datasheets of the FET's here:
http://michaelq.home.xs4all.nl/Datasheets/2SJ313.pdf
http://michaelq.home.xs4all.nl/Datasheets/2SK2013.pdf
http://michaelq.home.xs4all.nl/Datasheets/VN0210N5.pdf
http://michaelq.home.xs4all.nl/Datasheets/VP0210N5.pdf
Hi Xslavic,
Thanks for your reply. That seems contradictionairy, because they are much smaller than the 2SJ313 and 2SK2013's, so they can dissipate less. Which parameters are better than the 2SJ313 and 2SK2013's?
I am sorried, but I don't entirely know what you mean with: zeners power is relative to voltage voltage vs Watt.
What is the difference between a 1.3W zener and a 400mW zener? The power they can dissipate? Or the power they need?
Thanks for your reply. That seems contradictionairy, because they are much smaller than the 2SJ313 and 2SK2013's, so they can dissipate less. Which parameters are better than the 2SJ313 and 2SK2013's?
I am sorried, but I don't entirely know what you mean with: zeners power is relative to voltage voltage vs Watt.
What is the difference between a 1.3W zener and a 400mW zener? The power they can dissipate? Or the power they need?
Sorry for the unnecessary kick, I was not able to edit the post above to add the next text
I was able to change a few things the last few days to the build. There just were some aspects which I was not realy statisfied with. I've bought a few nice fuse sockets, so why not using them. I added a fuse in series with the primairies of the little transformer which powers the speaker protection. Chances of failure are not very likely, but better safe than sorried. I think I will add in the next few days also fuses in series with the secondairies of the 'big' transformers.
I also changed the connections to the TO-3 devices. At the time I mounted them to the heatsinks almost a year ago, I soldred thick copper wires directly to the leads of them. But I realised later that the copper wires could, when you apply little force, easily detach from the leads of the TO-3 devices. The warm / cold enviroment inside the amplifier does no good to the strength of the solder connections, so it will even get worse. In worst case scenario the copper wires could detach from the leads and make short circuit to the heatsink. Sure this would give nice fireworks, but hey 31 december is actually far away, and to be fair I prefer a working home made amplifier above home made fireworks haha. The copper wires were isolated by PVC. Altough PVC has a melting point at about 180 degrees Celsius, it's glass temperature (Tg, which is temperature at which the isolation becomes a bit brittle/liquid) is 80 degrees. Since the heatsinks get hot, no doubt about that, I had to think of a better candidate. So I replaced the cable with thick öilflex cable. This cable can withstand 180 degrees easily. I tried to lit it on fire (Kaplaars is sometimes a little pyromaniac haha), but that was very hard. Conclusion great cable for the task! I attached little thimbles to the cables. This way I made a much more reliable and stronger connection. Offcourse, the best solution would be using 'real' TO-3 sockets, but they would not fit perfectly to the heatsinks, so I hope this compromise will do.
Picture:
I was able to change a few things the last few days to the build. There just were some aspects which I was not realy statisfied with. I've bought a few nice fuse sockets, so why not using them. I added a fuse in series with the primairies of the little transformer which powers the speaker protection. Chances of failure are not very likely, but better safe than sorried. I think I will add in the next few days also fuses in series with the secondairies of the 'big' transformers.
An externally hosted image should be here but it was not working when we last tested it.
I also changed the connections to the TO-3 devices. At the time I mounted them to the heatsinks almost a year ago, I soldred thick copper wires directly to the leads of them. But I realised later that the copper wires could, when you apply little force, easily detach from the leads of the TO-3 devices. The warm / cold enviroment inside the amplifier does no good to the strength of the solder connections, so it will even get worse. In worst case scenario the copper wires could detach from the leads and make short circuit to the heatsink. Sure this would give nice fireworks, but hey 31 december is actually far away, and to be fair I prefer a working home made amplifier above home made fireworks haha. The copper wires were isolated by PVC. Altough PVC has a melting point at about 180 degrees Celsius, it's glass temperature (Tg, which is temperature at which the isolation becomes a bit brittle/liquid) is 80 degrees. Since the heatsinks get hot, no doubt about that, I had to think of a better candidate. So I replaced the cable with thick öilflex cable. This cable can withstand 180 degrees easily. I tried to lit it on fire (Kaplaars is sometimes a little pyromaniac haha), but that was very hard. Conclusion great cable for the task! I attached little thimbles to the cables. This way I made a much more reliable and stronger connection. Offcourse, the best solution would be using 'real' TO-3 sockets, but they would not fit perfectly to the heatsinks, so I hope this compromise will do.
Picture:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Wow it is realy quiet here. But not in Wageningen, Holland were a KSA-100 is born because, yes friends........ IT FINALLY WORKS!!!!!!!
I was realy nervous when powering it up. I've hooked the dummyload together with a Zobel network to the outputs. This was it, was al the effort I've made going to pay off, or did I make a nice piece of expensive firework. With a fire extinguisher in my left hand and with my right hand across the dial of the variac, I slowly powered it on. Hmm, no fireworks... that seems good. And It kept good!!!
But there is a but friends, I got no bias to one halve of the MJE15004's. I discovered a little mistake on Jims boards, there is missing a little track from the emitter of the middle MJE15032 to the MJE15033 compaired to the original. So I had no bias on connection DRIVE P1. But that was fixed easily with a little wire:
After fixing this, there was also current from DRIVE P1. Everything seemed to go OK so I pressed the proverbial gas pedal a bit more ;-). Freakin unbelievable how much power it has. I've measured 40 VAC max output across 8 Ohm load!!! I was able dial the DC offset to 0 mV very easy. Bias spread is about 10 mV across the different TO-3 devices. Not bad, I since I have not matched the devices.
After some testing (it was 2:00 p.m. already) it was time for some music! It's is unbelievable how it sounds, I try to be as objective as possible, but it is so dynamic, so much power. No hiss, no hum, nothing!!! I've set bias to 450 mV right now. Fans are running full speed right now (fan control is not finished yet). For Dutch standards today is pretty tropical, it is about 26 degrees outside and here in house it is almost 33 degrees. But despite of that, the tunnel cooler gets mildly warm.
Glad that it works, and VERY VERY happy that it sounds even better!!!!!!! But there is still a lot of work to be done, because now I know it works I can and thus still have to mount the lot in the cabinet
I was realy nervous when powering it up. I've hooked the dummyload together with a Zobel network to the outputs. This was it, was al the effort I've made going to pay off, or did I make a nice piece of expensive firework. With a fire extinguisher in my left hand and with my right hand across the dial of the variac, I slowly powered it on. Hmm, no fireworks... that seems good. And It kept good!!!
But there is a but friends, I got no bias to one halve of the MJE15004's. I discovered a little mistake on Jims boards, there is missing a little track from the emitter of the middle MJE15032 to the MJE15033 compaired to the original. So I had no bias on connection DRIVE P1. But that was fixed easily with a little wire:
After fixing this, there was also current from DRIVE P1. Everything seemed to go OK so I pressed the proverbial gas pedal a bit more ;-). Freakin unbelievable how much power it has. I've measured 40 VAC max output across 8 Ohm load!!! I was able dial the DC offset to 0 mV very easy. Bias spread is about 10 mV across the different TO-3 devices. Not bad, I since I have not matched the devices
.
After some testing (it was 2:00 p.m. already) it was time for some music! It's is unbelievable how it sounds, I try to be as objective as possible, but it is so dynamic, so much power. No hiss, no hum, nothing!!! I've set bias to 450 mV right now. Fans are running full speed right now (fan control is not finished yet). For Dutch standards today is pretty tropical, it is about 26 degrees outside and here in house it is almost 33 degrees. But despite of that, the tunnel cooler gets mildly warm.
Glad that it works, and VERY VERY happy that it sounds even better!!!!!!! But there is still a lot of work to be done, because now I know it works I can and thus still have to mount the lot in the cabinet
Last edited:
I am also VERY VERY happy that you stayed on this thread. Where else would I have found the little missing emitter track. Thank you so much. Now I am wondering what speaker system you will setup for this successful power amplifier. No hiss and no hum. Will you still change the 75C temperture sensor to a higher sensor? Very nice dummy load--May I ask what was the current to it at full load? Are you expecting more than 100 watts?Wow it is realy quiet here. But not in Wageningen, Holland were a KSA-100 is born because, yes friends........ IT FINALLY WORKS!!!!!!!
I was realy nervous when powering it up. I've hooked the dummyload together with a Zobel network to the outputs. This was it, was al the effort I've made going to pay off, or did I make a nice piece of expensive firework. With a fire extinguisher in my left hand and with my right hand across the dial of the variac, I slowly powered it on. Hmm, no fireworks... that seems good. And It kept good!!!
An externally hosted image should be here but it was not working when we last tested it.
But there is a but friends, I got no bias to one halve of the MJE15004's. I discovered a little mistake on Jims boards, there is missing a little track from the emitter of the middle MJE15032 to the MJE15033 compaired to the original. So I had no bias on connection DRIVE P1. But that was fixed easily with a little wire:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
After fixing this, there was also current from DRIVE P1. Everything seemed to go OK so I pressed the proverbial gas pedal a bit more ;-). Freakin unbelievable how much power it has. I've measured 40 VAC max output across 8 Ohm load!!! I was able dial the DC offset to 0 mV very easy. Bias spread is about 10 mV across the different TO-3 devices. Not bad, I since I have not matched the devices
An externally hosted image should be here but it was not working when we last tested it.
After some testing (it was 2:00 p.m. already) it was time for some music! It's is unbelievable how it sounds, I try to be as objective as possible, but it is so dynamic, so much power. No hiss, no hum, nothing!!! I've set bias to 450 mV right now. Fans are running full speed right now (fan control is not finished yet). For Dutch standards today is pretty tropical, it is about 26 degrees outside and here in house it is almost 33 degrees. But despite of that, the tunnel cooler gets mildly warm.
Glad that it works, and VERY VERY happy that it sounds even better!!!!!!! But there is still a lot of work to be done, because now I know it works I can and thus still have to mount the lot in the cabinet
Feels good after all the hard work doesn't it? Glad to hear you kept the smoke contained.
What does a 10k squarewave look like? Nice and square corners, peaks, or any fuzz on top of the signal?
If you ever do a full load power test, you will be inputing a 1vrms sine with and without a 8 ohm load.
A full frequency sweep should show equal amplitude with no peaking
A severe test is at full load you remove a load and place a .1uf 400v film cap and remove the cap quickly if it oscillates. This must be a tight connection if you decide to do this. I used a banana plug with the cap across it.
This will cover most of worst speaker loads you might come across and bring any issues to the surface.
Regards
David
What does a 10k squarewave look like? Nice and square corners, peaks, or any fuzz on top of the signal?
If you ever do a full load power test, you will be inputing a 1vrms sine with and without a 8 ohm load.
A full frequency sweep should show equal amplitude with no peaking
A severe test is at full load you remove a load and place a .1uf 400v film cap and remove the cap quickly if it oscillates. This must be a tight connection if you decide to do this. I used a banana plug with the cap across it.
This will cover most of worst speaker loads you might come across and bring any issues to the surface.
Regards
David
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.