Just fired it up today. The voltage rails are around 18VDC so I was aiming close to 5 watts per channel. The reason I built this unit is because I had a hard time finding a low wattage rackmount amp to power a pair of nearfield studio monitors (Yamaha NS-10).
After looking at prices for a fully balanced rackmount amp (and almost all are into the 500watt + power range), I knew it would be FAR cheaper to build the SoZ. Also the majority of all rackmount power amps like QSC, Crown, Quest, etc. use big noisy fans which I can not accept when i'm recording music on my PC. My PC has already been moidified for fanless power supply and the CPU fan has been software throttled down. Now if I can find a way to make my hard drives more quiet I would.
Space was quite limited and I had to thoroughly think in 3D. Certain components had to go in order so that everything could be put together. The capacitors are quite close to the donuts and I thought it would create noise. However, I only found that the right channel seemed to produce a hum which is only audiable going right up to it. For some reason the left channel barely has any noise (have to put my ear right up against the grill in order to hear hum) - yet, i've made all component layouts for each channel balanced. It is also worthy to note that the right channel that has more hum has a DC offset of around 200mV while the left channel is only 50mV. I suppose better matching of the mosfets may help? Should I bother?
Another design feature that I chose is not to use power resistors (heatsink kind). Instead I went with the cheaper generic ceramic wirewound ones.
Comments are welcome!
After looking at prices for a fully balanced rackmount amp (and almost all are into the 500watt + power range), I knew it would be FAR cheaper to build the SoZ. Also the majority of all rackmount power amps like QSC, Crown, Quest, etc. use big noisy fans which I can not accept when i'm recording music on my PC. My PC has already been moidified for fanless power supply and the CPU fan has been software throttled down. Now if I can find a way to make my hard drives more quiet I would.
Space was quite limited and I had to thoroughly think in 3D. Certain components had to go in order so that everything could be put together. The capacitors are quite close to the donuts and I thought it would create noise. However, I only found that the right channel seemed to produce a hum which is only audiable going right up to it. For some reason the left channel barely has any noise (have to put my ear right up against the grill in order to hear hum) - yet, i've made all component layouts for each channel balanced. It is also worthy to note that the right channel that has more hum has a DC offset of around 200mV while the left channel is only 50mV. I suppose better matching of the mosfets may help? Should I bother?
Another design feature that I chose is not to use power resistors (heatsink kind). Instead I went with the cheaper generic ceramic wirewound ones.
Comments are welcome!
Attachments
OT - Hard disk noise
Super_BQ,
I have used a product called "SilentDrive" that encloses a 3.5 HD (and then fits into a 5.25 bay) and gives a worthwhile decrease in noise. My machine runs 24*7 (for about the last 4 months) and has not suffered overheating as a result (the SilentDrive incorperates an alloy heatsink). As long as your internal case temprature is not to high you should be OK. They are a little difficult to fit and they are not cheap but it keeps my 'other half happy
Dave
Super_BQ,
I have used a product called "SilentDrive" that encloses a 3.5 HD (and then fits into a 5.25 bay) and gives a worthwhile decrease in noise. My machine runs 24*7 (for about the last 4 months) and has not suffered overheating as a result (the SilentDrive incorperates an alloy heatsink). As long as your internal case temprature is not to high you should be OK. They are a little difficult to fit and they are not cheap but it keeps my 'other half happy
Dave
moe29,
I've extensively tested the resistors. For each 8 ohm bank are 4 x 33ohm 10 watt rated resistors in parallel. I've also matched all the resistors for an equal value in each bank. Do they get hot? At 18v, i'm just able to keep my finger on it. Using my Variac, i've operated the amp over 20v rails and the resistors do get more hot - but far from what I think as they would fail. Though, a lot of the heat does get dissipated on the rack where they are strapped down on. I really don't see why people don't use more of these resistors as they're cheap and don't need the heatsinking. I've seen ones that go up to 25watts (like 5" long) and would work no problem at 10w output SoZ. (non-inductive also available)
My other alternative for resistors is to use the Caddocks and heatsink them (TO-126, TO-220, etc). But I wanted to keep this project on a budget as it would not be used for high-end hifi purposes. However, I would probably be hard pressed to hear the difference. This amp does sound A LOT better (more dynamics!) than when I was driving my NS10 with that Korsun U2 (M.L. Rosette single ended int. amp).
As for the caps. they are:
Make: Sprague
Series: Powerlytic
Model#: 36D4233P
Rated: 54000uF - 30VDC
I bought these 2nd hard last year for only $4us each (locally at a surplus shop). I do have new snap-in electro type Panasonics that are 1/8th the size for even more uF and voltage but saving them for zen v4.
For bypass caps I used Dayton polypropylenes. 8.2uF 250VDC and 2.0uF 250VDC ones on each big can. Ideally 54uF and 5.4uF and .54uF but I just don't have the space. Not even room for inductors to do a pi-filter.
BQ
I've extensively tested the resistors. For each 8 ohm bank are 4 x 33ohm 10 watt rated resistors in parallel. I've also matched all the resistors for an equal value in each bank. Do they get hot? At 18v, i'm just able to keep my finger on it. Using my Variac, i've operated the amp over 20v rails and the resistors do get more hot - but far from what I think as they would fail. Though, a lot of the heat does get dissipated on the rack where they are strapped down on. I really don't see why people don't use more of these resistors as they're cheap and don't need the heatsinking. I've seen ones that go up to 25watts (like 5" long) and would work no problem at 10w output SoZ. (non-inductive also available)
My other alternative for resistors is to use the Caddocks and heatsink them (TO-126, TO-220, etc). But I wanted to keep this project on a budget as it would not be used for high-end hifi purposes. However, I would probably be hard pressed to hear the difference. This amp does sound A LOT better (more dynamics!) than when I was driving my NS10 with that Korsun U2 (M.L. Rosette single ended int. amp).
As for the caps. they are:
Make: Sprague
Series: Powerlytic
Model#: 36D4233P
Rated: 54000uF - 30VDC
I bought these 2nd hard last year for only $4us each (locally at a surplus shop). I do have new snap-in electro type Panasonics that are 1/8th the size for even more uF and voltage but saving them for zen v4.
For bypass caps I used Dayton polypropylenes. 8.2uF 250VDC and 2.0uF 250VDC ones on each big can. Ideally 54uF and 5.4uF and .54uF but I just don't have the space. Not even room for inductors to do a pi-filter.
BQ
I guess spreading the load with multiple resistors
helps with heat dissipation too. I think Mr. Pass
mentioned that you can run resistors closer to their
limits than MOSFETS.
what's cool about it is that you had an idea and
gave it a try, i guess that's what DIY is all about.
... i was wondering about inductors... it is hard to
keep the SOZ compact! : )
thanks,
moe29
helps with heat dissipation too. I think Mr. Pass
mentioned that you can run resistors closer to their
limits than MOSFETS.
what's cool about it is that you had an idea and
gave it a try, i guess that's what DIY is all about.
... i was wondering about inductors... it is hard to
keep the SOZ compact! : )
thanks,
moe29
bob12345678,
A chain store called Dick Smith (like Radio Shack) is where I bought the case from.
http://www.dse.co.nz/cgi-bin/dse.storefront/3e2deb76052470f42741c0a87f990716/Product/View/H2481
Works out to be under $50usd. If I figure the time and labour to build a similar case, it would cost more. What ever drilling and cutting I had to do was quite minimal.
BQ
A chain store called Dick Smith (like Radio Shack) is where I bought the case from.
http://www.dse.co.nz/cgi-bin/dse.storefront/3e2deb76052470f42741c0a87f990716/Product/View/H2481
Works out to be under $50usd. If I figure the time and labour to build a similar case, it would cost more. What ever drilling and cutting I had to do was quite minimal.
BQ
The Kinda... Funny Looking Guy
Mr. Pass,
You guessed right. I found his role in Reservoir Dogs and Fargo was right on!
Also forgot to mention that I used IRFP044 mosfets.
http://www.irf.com/product-info/datasheets/data/irfp044.pdf
BQ
Mr. Pass,
You guessed right. I found his role in Reservoir Dogs and Fargo was right on!
Also forgot to mention that I used IRFP044 mosfets.
http://www.irf.com/product-info/datasheets/data/irfp044.pdf
BQ
Hum Discovery
For an update. What ever small hum I have in the right channel may be due to the capacitors. I've kept the design of the amp completely symetrical. The AC power wires run down the center so that wouldn't be causing the hum.
So as for using 2nd hand caps - don't expect them to run like new.
The amp has been going for 100s of hours and no problems at all. The concern with the tie straps melting hasn't been a concern. Maybe those resistors just simply don't get hot enough . Though with the way they are posistioned ; if they do melt, the resistors will sit flat rather than hang dangling around.
BQ
Regarding about my PC and noisy fans. Well unfortunately the motherboard packed up. My friend insisited that the motherboard needed air circulation which caused it to fail. The software diagnostics I was using showed the MB temperature was around 40°. This time i've bought a replacement board and underclocked the CPU to run only at 1.8Ghz than from 2.4Ghz ; less power draw from CPU = less heat generated inside from the heatsinks.
http://www.geocities.com/super_bq/Misc/IMG_1019.jpg
For an update. What ever small hum I have in the right channel may be due to the capacitors. I've kept the design of the amp completely symetrical. The AC power wires run down the center so that wouldn't be causing the hum.
So as for using 2nd hand caps - don't expect them to run like new.
The amp has been going for 100s of hours and no problems at all. The concern with the tie straps melting hasn't been a concern. Maybe those resistors just simply don't get hot enough . Though with the way they are posistioned ; if they do melt, the resistors will sit flat rather than hang dangling around.
BQ
Regarding about my PC and noisy fans. Well unfortunately the motherboard packed up. My friend insisited that the motherboard needed air circulation which caused it to fail. The software diagnostics I was using showed the MB temperature was around 40°. This time i've bought a replacement board and underclocked the CPU to run only at 1.8Ghz than from 2.4Ghz ; less power draw from CPU = less heat generated inside from the heatsinks.
http://www.geocities.com/super_bq/Misc/IMG_1019.jpg
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.