Hi David,
Thanks for your input! Hmm, that gives food for thought. My motivation behind mounting the tunnel coolers together and place them like post 32 was to get better air circularion since the air can flow directly in and out the cabinet through the tunnel.
Basicly there are two fans, one which sucks air from outside and blows air through the tunnel. At the end of the tunnel I mount another fan which sucks the air out of the tunnel and blows it out of the cabinet (not in the picture). That would fit just right from side to side. Therefore I have to drill two holes in the cabinet at each side, so air can flow freely. When I mounted the tunnel coolers as in post 24 there would be less circulation of air was the idea. On the other side, in the new 'mounted' configuration differences in temperature are indeed almost inevitable :-(
In the original Krell configuration the tunnel coolers were placed upright, but I find that the least best option. It would give two ugly holes on top and two at the bottom for optimal airflow.
I think that it is a good idea to experiment a bit with the best configuration before I drill large holes in the cabinet.
"if" I had to use fans, I would use just one at the end of tunnel (one fan= less noise) it will pull the hot air trapped inside tunnel out.
Hmm that would also be an option. Although, the tunnel cooler would in that case not fit precisely from one side to the other in the cabinet.
To be fair there is not very much noise from the fans. Noise is reduced a lot when I run the fans with less 'juice'. For example, when I run them at half speed they are almost inaudible, but still a lot of airflow. I am thinking about some kind of fancontrol which moderates fan speed with temperature. But first find some time to realise this
To be fair there is not very much noise from the fans. Noise is reduced a lot when I run the fans with less 'juice'. For example, when I run them at half speed they are almost inaudible, but still a lot of airflow. I am thinking about some kind of fancontrol which moderates fan speed with temperature. But first find some time to realise this
Last edited:
You don,t need alot of fan speed so the idea of pulling air around the toroids and caps is a moot point. It does make a difference however if you are using more than say 1200 rpm. What becomes an issue with high fan speeds is noise due to close proximity of the incoming air.
Your last picture is great for low noise since their is no objects to impede incoming air but bad for even heat distribution
So looking at how Krell did theirs was the best overall configuration for both fan noise and even heat in the available enclosure.
Their orientation was the best overall compromise for both considerations
I like to keep fan speeds under 1000 rpm if I can,t meet the space requirements. And that is very close to the turn on voltage for DC fans.
It becomes really, a first and paramount design requiement when doing a layout for the chasiss
Regards
David
Your last picture is great for low noise since their is no objects to impede incoming air but bad for even heat distribution
So looking at how Krell did theirs was the best overall configuration for both fan noise and even heat in the available enclosure.
Their orientation was the best overall compromise for both considerations
I like to keep fan speeds under 1000 rpm if I can,t meet the space requirements. And that is very close to the turn on voltage for DC fans.
It becomes really, a first and paramount design requiement when doing a layout for the chasiss
Regards
David
Andrew, Andrew, Andrew... what is your problem. You keep nagging that it is a bad idea to place NTC's parallel. OK maybe, but why? Is it that hard to underpin your theorem? I don't know.... but you keep refusing to give a constructive answer. It is like I state driving on 3 wheels is a bad idea, but I don't say why it is a bad idea. A critic dutchman asks why, but all I say, read the English. What English, were is the explanation? Luckaly CBS240 explained it, so it is no longer necessairy.
Other thing, you state that bypassing the rectifier diodes with capacitors is a bad idea. Again, OK, can be, but why? Still got no answer from you. Again CBS240 was the helping friend. Is it that strange to ask why?
But Andrew, lets not spoil this threat with our little discussion. To be fair my initial question had nothing to do with the NTC's. My initial question was about the series capacitor. The NTC's are NOT really my problem, and to be fair this config works perfect, it switches the two 1 KVA toroids perfectly on, and no fuse is spoiled.
Other thing, you state that bypassing the rectifier diodes with capacitors is a bad idea. Again, OK, can be, but why? Still got no answer from you. Again CBS240 was the helping friend. Is it that strange to ask why?
But Andrew, lets not spoil this threat with our little discussion. To be fair my initial question had nothing to do with the NTC's. My initial question was about the series capacitor. The NTC's are NOT really my problem, and to be fair this config works perfect, it switches the two 1 KVA toroids perfectly on, and no fuse is spoiled.
Hi Friends,
Little update about the series capacitor were I did not know the meaning of. Learned from a Dutch forum that it does no harm to the amplifier When the capacitor makes short circuit the amplifier just turns on, so no problem there (I should have know that! ).
The amplifier is always in a sort of standby due to this capacitor. But with very little current. In ideal circumstances the current in stand by is: L = 2.Pi.U.f.c, so in my case that is 2*pi*230*50*0,000000022 = 1,6mA. That explains why the current drops almost to zero when I put a little load after the soft start.
The function of the capacitor is to damp little sparks which occur when the relays disconnect (Lentz law).
Little update about the series capacitor were I did not know the meaning of. Learned from a Dutch forum that it does no harm to the amplifier
When the capacitor makes short circuit the amplifier just turns on, so no problem there (I should have know that! ).The amplifier is always in a sort of standby due to this capacitor. But with very little current. In ideal circumstances the current in stand by is: L = 2.Pi.U.f.c, so in my case that is 2*pi*230*50*0,000000022 = 1,6mA. That explains why the current drops almost to zero when I put a little load after the soft start.
The function of the capacitor is to damp little sparks which occur when the relays disconnect (Lentz law).
AVWERK is correct, Kaplaars.
the last configuration you've shown will mean you have one channel running hot, the one furthest from the fan pushing air through the tunnel.
The air entering this channel will already be hot from running through the first half of your tunnel.
Groet,
Klaas
Hi Klaas,
Hmm, I hope the temperature does not get to hot. Aesthetically I think that this is the best looking configuration. I got inspired by A D&B amplifier. David & Ben used this concept comercially. They made beasts of Class A amplifiers, which I think are even more beautifull than Krell. This is an example:
http://michaelq.home.xs4all.nl/Projects/D&B.jpg
Fortunately the tunnel coolers are easy to separate, so if temperature gets out of control it is not very hard to choose a different configuration.
By the way, I added extra capacitors parallel to the big electrolytic caps. The capacitors are a bit overkill, but I have a lot of these in stock. The advantages with this caps overweight potential disadvantages, but if it turns out that they do more bad than good I can easily remove them.
Hmm, I hope the temperature does not get to hot. Aesthetically I think that this is the best looking configuration. I got inspired by A D&B amplifier. David & Ben used this concept comercially. They made beasts of Class A amplifiers, which I think are even more beautifull than Krell. This is an example:
http://michaelq.home.xs4all.nl/Projects/D&B.jpg
An externally hosted image should be here but it was not working when we last tested it.
Fortunately the tunnel coolers are easy to separate, so if temperature gets out of control it is not very hard to choose a different configuration.
By the way, I added extra capacitors parallel to the big electrolytic caps. The capacitors are a bit overkill, but I have a lot of these in stock. The advantages with this caps overweight potential disadvantages, but if it turns out that they do more bad than good I can easily remove them.
An externally hosted image should be here but it was not working when we last tested it.
Connecting 2 NTC resistors in parellel will almost certainly result in 1 being hot and having a low resistance while the other is cold and having a high resistance. Clearly one of them will end up carrying almost all the current and, quite possibly, be outside it's safe operating area (which is likely why you didn't just use one in the first place).
dc
dc
Aha so that is why it is not smart to set NTC's in parallel! Did not think of that, I forgot about the R/T link which is offcourse not exactely the same for each NTC; one can heat faster than the other. That would be the explanation, thank you DRC, mystery solved, learned another lesson
So actually, when I have two parallel sets in series there are in worst case scenario always two NTC's in series were most current goes through (one in the first parallel set which heats up first and therefore conducting most current and one in the other parallel set which does the same). The other two NTC's which are parallel to the 'warm' NTC's conduct only little current due to their high resistance relative to the other two.
Hmm, gives food for thought, off course the NTC's are pretty close to each other, so they will heat each other also a bit, but it is obviously not an ideal configuration the way it is now.
Edit: typo
So actually, when I have two parallel sets in series there are in worst case scenario always two NTC's in series were most current goes through (one in the first parallel set which heats up first and therefore conducting most current and one in the other parallel set which does the same). The other two NTC's which are parallel to the 'warm' NTC's conduct only little current due to their high resistance relative to the other two.
Hmm, gives food for thought, off course the NTC's are pretty close to each other, so they will heat each other also a bit, but it is obviously not an ideal configuration the way it is now.
Edit: typo
Last edited:
I just wanted to congratulate you for building this Amp. As much as I try to learn electronics almost all of it goes over my head.
I have a Krell KSA150 and it is a wonderful amp but scares the **** out of me. Even just touching it when it is on kinda scares me knowing how much current it draws.
Good luck I hope it all works out as it should.
I have a Krell KSA150 and it is a wonderful amp but scares the **** out of me. Even just touching it when it is on kinda scares me knowing how much current it draws.
Good luck I hope it all works out as it should.
Kaplaars
Your build quality is excellent ! Rich student?
If you in live a cold area you might just get away with your present layout. However I would suggest that you consider adding a shutdown mechanism should you experience a fan failure or overheating problem.
BTW: Subjectively I consider the KSA150 to be exceptional in dynamics but the KSA100 has a better 3 dimensional soundstage.
Regards
Jozua
Your build quality is excellent ! Rich student?
If you in live a cold area you might just get away with your present layout. However I would suggest that you consider adding a shutdown mechanism should you experience a fan failure or overheating problem.
BTW: Subjectively I consider the KSA150 to be exceptional in dynamics but the KSA100 has a better 3 dimensional soundstage.
Regards
Jozua
Last edited:
I'd arrange the tunnels to suck air from in the front panel (with a filter at the front panel) past the power supply and then thru the fans blowing into the tunnels and out the back. Fans only at the INSIDE ends of the tunnels to reduce noise at the outside. But that's 'cause I sometimes rack my amps and want the heat to go out the back rather than heating the sides of other equipment in the rack. The relatively large volume of air will not increase in temp much from flowing around the tranny and filter caps before the tunnels, but will make the filter caps last a LOT longer from not being exposed to the tranny and rectifier heat. Also, if the air is drawn from inside the cabinet, if the tunnels ever become really warm they'll start radiating to the outside of the tunnels too, and you will really appreciate the fans evacuating that hot air. In fact, if you think about it, the tunnels and rows of transistors probalby WILL be radiating quite a bit of heat into the cabinet if the fans are slow & quiet.
If you leave it one tunnel cross-flow, with no forced cooling for the power supply, are you going to add ventilation holes for the power supply? If so, place the air holes carefully so that the power supply is not taking in its own fan exhaust if you put it in a big rack cabinet! If you're sucking in on one side and blowing hot out the other side, you can't have intake on either side, top, or bottom...pretty much leaving just the front, no cross-flow and little natural chimney convection.
If you fan-cool only inside the tunnels, without drawing air from inside the cab, it is nice that the power supply and circuitry will stay relatively cleaner. My amps use noisy fans on the cabinet back that draw air thru everything, and the insides of the amps get filthy without intake filters.
I would also probably mount the rectifiers directly to the cabinet, just to get half the heat out from the start.
If you leave it one tunnel cross-flow, with no forced cooling for the power supply, are you going to add ventilation holes for the power supply? If so, place the air holes carefully so that the power supply is not taking in its own fan exhaust if you put it in a big rack cabinet! If you're sucking in on one side and blowing hot out the other side, you can't have intake on either side, top, or bottom...pretty much leaving just the front, no cross-flow and little natural chimney convection.
If you fan-cool only inside the tunnels, without drawing air from inside the cab, it is nice that the power supply and circuitry will stay relatively cleaner. My amps use noisy fans on the cabinet back that draw air thru everything, and the insides of the amps get filthy without intake filters.
I would also probably mount the rectifiers directly to the cabinet, just to get half the heat out from the start.
Last edited:
Your config has a fan at each end of the double-tunnel, where they can be heard the most from outside the cabinet. One of the nicest things about such tunnels is that you can bury the fans inside, and even add a muffler plate to the outside if necessary. Check out the nearly-silent z mainframe cabinets; one large sound muffler.
Hi Jozua,
Thanks! Hahah I would wish! ;-) But unfortunatelly that is not the case, still have not won the national lotery :-(
But to be fair having little budget is in some way a good thing. When you have little budget it forces you to be creative. For me it is fun and a kind of sport to obtain parts for as least money as possible (no, with that sentence I do not mean stealing!! haha). So I search a lot on Ebay and especially Marktplaats (Marktplaats is a Dutch equivalent of Ebay) and try to buy parts as cheap as possible.
I've bought the cabinet, the softstart and the speaker protection PCB's new. The cabinet was the most expensive part and costed me about 125 euro's. The rest are actually second hand parts, and costed me thusfar together about 200-250 euro's. For me still pretty much money, but from the other hand pretty cheap when I compare it to what others spended. I think if I had to buy everything brand new the counter would be at 700 euro's.
The climate here is pretty temperate, so not too cold not too warm. I have thought about it and I think I will use the Krell clone especially during the winter and the AudioAnalyse below during the summer:
Haha only joking. The AudioAnalyse is still 60W class A. A real heat generating monster, but love it's sound. Can't almost imagine how much heat the Krell generates since the Krell is fully class A when driving it at 8 ohm load.
That is a very good suggestion Jozua, I have build in a form of thermal protection. I have two TO220 thermostats. On the softstart PCB there is a little circuitry were the thermostats are connected to. When the temperature at the tunnel coolers exceed 75 degrees Celsius and the thermostats sense this, the relays at the softstart shut down and thus the amplifier itself shuts down.
I am doubting if I have to buy thermostats which can handle higher temperature. When I compare to the AudioAnalyse 70 degrees heatsinks are not uncommon.
@Cyclecamper. Thanks for your reply! Hmm I think it would be a bit of a shame to cut a large hole in the front panel. I like the look how the frontpanel is very much. The frontpanel is made from 10mm thick steel, so very difficult do cut through. The amplifier will have natural air circulation around it because I don't want to put it in a cabinet but set it on the floor instead (I almost have to, it weights about 50kg allready). Radiation is indeed a good point! I think radiation from the tunnel cooler is almost inevitable. Maybe I place an additional 12 of 14cm fan on top of the cabinet.
On top of the cabinet there are ventilation holes, but not above the PSU:
I hope the copper bars are sufficient enough keep the bridge rectifiers cool (the Cu-bars are about 8mm thick). If they are not, placing the bridges directly to the cabinet is a good solution, will remember that.
I've bought fanfilters. They match the cabinet very well and are almost the same as shown in the D&B amplifier earlier These ones: http://www.nexustek.nl/NXS-nexus-ff-series-fan-filters.htm
I will also add four of these silicone absorbers: http://www.nexustek.nl/NXS-nexus-sa-series-silicone-fan-noise-absorber.htm
Edit: reaction to Cyclecamper
Thanks! Hahah I would wish! ;-) But unfortunatelly that is not the case, still have not won the national lotery :-(
But to be fair having little budget is in some way a good thing. When you have little budget it forces you to be creative. For me it is fun and a kind of sport to obtain parts for as least money as possible (no, with that sentence I do not mean stealing!! haha). So I search a lot on Ebay and especially Marktplaats (Marktplaats is a Dutch equivalent of Ebay) and try to buy parts as cheap as possible.
I've bought the cabinet, the softstart and the speaker protection PCB's new. The cabinet was the most expensive part and costed me about 125 euro's. The rest are actually second hand parts, and costed me thusfar together about 200-250 euro's. For me still pretty much money, but from the other hand pretty cheap when I compare it to what others spended. I think if I had to buy everything brand new the counter would be at 700 euro's.
The climate here is pretty temperate, so not too cold not too warm. I have thought about it and I think I will use the Krell clone especially during the winter and the AudioAnalyse below during the summer:
An externally hosted image should be here but it was not working when we last tested it.
Haha only joking. The AudioAnalyse is still 60W class A. A real heat generating monster, but love it's sound. Can't almost imagine how much heat the Krell generates since the Krell is fully class A when driving it at 8 ohm load.
That is a very good suggestion Jozua, I have build in a form of thermal protection. I have two TO220 thermostats. On the softstart PCB there is a little circuitry were the thermostats are connected to. When the temperature at the tunnel coolers exceed 75 degrees Celsius and the thermostats sense this, the relays at the softstart shut down and thus the amplifier itself shuts down.
I am doubting if I have to buy thermostats which can handle higher temperature. When I compare to the AudioAnalyse 70 degrees heatsinks are not uncommon.
@Cyclecamper. Thanks for your reply! Hmm I think it would be a bit of a shame to cut a large hole in the front panel. I like the look how the frontpanel is very much. The frontpanel is made from 10mm thick steel, so very difficult do cut through. The amplifier will have natural air circulation around it because I don't want to put it in a cabinet but set it on the floor instead (I almost have to, it weights about 50kg allready). Radiation is indeed a good point! I think radiation from the tunnel cooler is almost inevitable. Maybe I place an additional 12 of 14cm fan on top of the cabinet.
On top of the cabinet there are ventilation holes, but not above the PSU:
An externally hosted image should be here but it was not working when we last tested it.
I hope the copper bars are sufficient enough keep the bridge rectifiers cool (the Cu-bars are about 8mm thick). If they are not, placing the bridges directly to the cabinet is a good solution, will remember that.
I've bought fanfilters. They match the cabinet very well and are almost the same as shown in the D&B amplifier earlier
These ones: http://www.nexustek.nl/NXS-nexus-ff-series-fan-filters.htmI will also add four of these silicone absorbers: http://www.nexustek.nl/NXS-nexus-sa-series-silicone-fan-noise-absorber.htm
Edit: reaction to Cyclecamper
Last edited:
You could place the fans on the outside with each blowing toward each other. This would probably cancel all sound coming from the fans.
You would then have to seal off the transformer side and make one rectangular slot in the center rear so all positive pressure exits there.
I know you want it to look nice, but performance sometimes has its cost.
Regards
David
You would then have to seal off the transformer side and make one rectangular slot in the center rear so all positive pressure exits there.
I know you want it to look nice, but performance sometimes has its cost.
Regards
David
Little update I want to be as original as possible, so I've designed a back- and frontplate for the amplifier in Adobe Ilustrator. I sended the design to a company which will engrave the frontplate design in a gold coated plate and prints the backplate design on aluminium. Hope the design turns out realy nice on metal.
The basic design is as follows for the frontplate (almost the same as in the original):
And for the backplate:
Taped a paper example to the front:
I want to be as original as possible, so I've designed a back- and frontplate for the amplifier in Adobe Ilustrator. I sended the design to a company which will engrave the frontplate design in a gold coated plate and prints the backplate design on aluminium. Hope the design turns out realy nice on metal. The basic design is as follows for the frontplate (almost the same as in the original):
An externally hosted image should be here but it was not working when we last tested it.
And for the backplate:
An externally hosted image should be here but it was not working when we last tested it.
Taped a paper example to the front:
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.