This is what i dig out from my drawer last weekend.
9 - 0 - 9 2A transformer
a few parts enough for a single channel simplified F5
4 2200uF
0.47ohm 5watt
I know 12Vdc for F5 is too low, but i would like to give it a try, the only problem i have is, when the DC offset set to near 0V, the speaker humm, but when i set to -ve 600v offset above, the speaker will stop humming, what actually is it?
and lastly, it really can sing well @ 12V!
9 - 0 - 9 2A transformer
a few parts enough for a single channel simplified F5
4 2200uF
0.47ohm 5watt
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.
An externally hosted image should be here but it was not working when we last tested it.
I know 12Vdc for F5 is too low, but i would like to give it a try, the only problem i have is, when the DC offset set to near 0V, the speaker humm, but when i set to -ve 600v offset above, the speaker will stop humming, what actually is it?
and lastly, it really can sing well @ 12V!
when the DC offset set to near 0V, the speaker humm
but when i set to -ve 600v offset above, the speaker will stop humming
what actually is it?
you do not adjust DC, as such
you adjust bias balance
dc is the result of that
Regarding my cylinder head build:
I think the problem I had with R12 smoking when I hooked up the quads, were just to do with the speakers. I've had it on for 2 extended sessions hooked up to my horns with no issues. I have yet to go back and investigate more with the stacked quads.
I can say with more certainty now that while using the cylinder head works, it defintely doesn't dissipate the heat as well as the finned heatsink. To make it workable, I reinstalled the temp compensation circuitry - this keeps a check on it all and prevents any kind of runaway. Also, anyone thinking of trying this should know that the head gets to 60-63 degC (which is 40 above ambient) when in a shelf with poorish circulation. On my build, the actual fets go to the mid 80's temp wise. Still well within their ratings, but it feels hot.
Sound is good though!! This one was built with the fairchild 19P and 12N devices, everything else pretty much standard. I think it sounds a little more polite than the 9240/240 build, and the midrange might be a little more lean.... but still very good. I reckon this would sound very good with a good valve preamp - I ahve an aikido which I intend trying out. I have been using it with a DCB1 and I think the extreme neutrality of the DCB1 could do with a little more bloom.
Of course the amp has a total of about 10 hours on it - and after another 10 I'll probably go back and tweak it again just to check her out fully. Anyway, hope that helps the frugal!!!
Fran
I think the problem I had with R12 smoking when I hooked up the quads, were just to do with the speakers. I've had it on for 2 extended sessions hooked up to my horns with no issues. I have yet to go back and investigate more with the stacked quads.
I can say with more certainty now that while using the cylinder head works, it defintely doesn't dissipate the heat as well as the finned heatsink. To make it workable, I reinstalled the temp compensation circuitry - this keeps a check on it all and prevents any kind of runaway. Also, anyone thinking of trying this should know that the head gets to 60-63 degC (which is 40 above ambient) when in a shelf with poorish circulation. On my build, the actual fets go to the mid 80's temp wise. Still well within their ratings, but it feels hot.
Sound is good though!! This one was built with the fairchild 19P and 12N devices, everything else pretty much standard. I think it sounds a little more polite than the 9240/240 build, and the midrange might be a little more lean.... but still very good. I reckon this would sound very good with a good valve preamp - I ahve an aikido which I intend trying out. I have been using it with a DCB1 and I think the extreme neutrality of the DCB1 could do with a little more bloom.
Of course the amp has a total of about 10 hours on it - and after another 10 I'll probably go back and tweak it again just to check her out fully. Anyway, hope that helps the frugal!!!
Fran
My F5 is nearing completion.
I'll be shopping for speakers for it. Is there a general rule that I should be aware of when shopping for speakers for the F5.
I know speaker sensitivity is something to consider. I think I'll get some bookshelf speakers for it.
I was looking at the Zaph audio ZD5.
Any suggestions?
Thanks,
JG
I'll be shopping for speakers for it. Is there a general rule that I should be aware of when shopping for speakers for the F5.
I know speaker sensitivity is something to consider. I think I'll get some bookshelf speakers for it.
I was looking at the Zaph audio ZD5.
Any suggestions?
Thanks,
JG
F5 tempco thermistor issue
Has anyone experienced problems with gluing the Th1/Th2 thermistors to the drain terminals of the output mosfets, as is done in Nelson's F5 as shown in the 6moons.com review. (I cannot tell for sure that the thermistor is actually glued to the drain in the photos). I used a drop of Arctic Silver thermal adhesive to attach my thermistors and experienced the problem described below.
I finally got all the pieces together for my stereo F5 amplifier. The F5 channels were assembled using the cviller V1.0 F5 boards from the DIYAUDIO Store and the tech-diy.com Fairchild F5 parts.
Today I do the SMOKE TEST.
In the following adjustments, 8 ohm, 50 watt load resistors were connected to the outputs, and a variac was used to gradually ramp-up the AC voltage, and inputs were shorted to ground.
I started the "smoke test" with only the right channel powered. The P1 and P2 bias adjustments went according to the Nelson Pass F5 instructions with R11/R12 voltages of .59 V and an output offset voltage of about 10 mv. I applied a .8 volt P-P 1 KHz sine wave input and looked at the output using a Tek 465 scope. Looked fine.
I connected power to the left channel (leaving the right channel also powered) and began bias adjustments. The P1/P2 adjustments DID NOT behave in the same manner as for the right channel. I could not get the offset voltage to near zero. The offset was in the neighborhood of +150mv.
I powered everything off and did a thorough visual inspection to make sure all components were correctly installed and correctly soldered. I could not find any problems. I applied power again and looked at the output in the scope: it looked very bad, but since I did not photograph the scope, I would have to sketch the waveform from memory.
After a while in the adjustment process, the offset voltage would randomly alternate between +150mv and -100mv. Eventually, IT HAPPENED: SMOKE and the R11/R12 source resistors started glowing. A sure sign that the mosfets had become dead shorts. I removed the left channel board from the heatsink and discovered that the Th1 thermistor had become very hot, probably glowing also.
My theory is that a low resistance or short occured between the drain of Q3 and the node between Th1 and R15. This placed approximately 24 volts across Th1, causing it to heat up and reduce its resistance causing it to heat further into to thermal runaway. Spice simulations of the F5 show that a properly adjusted F5 channel will have problems if there is such a short: approximately 100W dissipated in Q3 and Q4. The problem could have been due to some of the thermal adhesive bridging between terminals of the thermistor and the drain of Q3, or a scratch in the coating of the thermistor itself, allowing direct contact to the drain.
Has anyone had a similar problem?
Has anyone experienced problems with gluing the Th1/Th2 thermistors to the drain terminals of the output mosfets, as is done in Nelson's F5 as shown in the 6moons.com review. (I cannot tell for sure that the thermistor is actually glued to the drain in the photos). I used a drop of Arctic Silver thermal adhesive to attach my thermistors and experienced the problem described below.
I finally got all the pieces together for my stereo F5 amplifier. The F5 channels were assembled using the cviller V1.0 F5 boards from the DIYAUDIO Store and the tech-diy.com Fairchild F5 parts.
Today I do the SMOKE TEST.
In the following adjustments, 8 ohm, 50 watt load resistors were connected to the outputs, and a variac was used to gradually ramp-up the AC voltage, and inputs were shorted to ground.
I started the "smoke test" with only the right channel powered. The P1 and P2 bias adjustments went according to the Nelson Pass F5 instructions with R11/R12 voltages of .59 V and an output offset voltage of about 10 mv. I applied a .8 volt P-P 1 KHz sine wave input and looked at the output using a Tek 465 scope. Looked fine.
I connected power to the left channel (leaving the right channel also powered) and began bias adjustments. The P1/P2 adjustments DID NOT behave in the same manner as for the right channel. I could not get the offset voltage to near zero. The offset was in the neighborhood of +150mv.
I powered everything off and did a thorough visual inspection to make sure all components were correctly installed and correctly soldered. I could not find any problems. I applied power again and looked at the output in the scope: it looked very bad, but since I did not photograph the scope, I would have to sketch the waveform from memory.
After a while in the adjustment process, the offset voltage would randomly alternate between +150mv and -100mv. Eventually, IT HAPPENED: SMOKE and the R11/R12 source resistors started glowing. A sure sign that the mosfets had become dead shorts. I removed the left channel board from the heatsink and discovered that the Th1 thermistor had become very hot, probably glowing also.
My theory is that a low resistance or short occured between the drain of Q3 and the node between Th1 and R15. This placed approximately 24 volts across Th1, causing it to heat up and reduce its resistance causing it to heat further into to thermal runaway. Spice simulations of the F5 show that a properly adjusted F5 channel will have problems if there is such a short: approximately 100W dissipated in Q3 and Q4. The problem could have been due to some of the thermal adhesive bridging between terminals of the thermistor and the drain of Q3, or a scratch in the coating of the thermistor itself, allowing direct contact to the drain.
Has anyone had a similar problem?
a bit small
CA18RLY/22TAF-G
dont know if it will be a problem having porting on back of speaker, in a boolshelf
my bools are usually tricky for shelf positioning
even when they're small , still hornless
So for the time being I decided that this is a non-issue, and will live happily until I check the bias again in few weeks. I made a quick check with my limited spectrum analyzer, and I do not really have the same figures that NP has, the second harmonic is still dominant, and distortion is higher on the right channel.
Just to be sure I understand how things works: is it correct to say that the DC offset is the effect of the bias difference between the two mosfet ? If I have 0.6 V on one side and zero offset, can I assume that the other mosfet is also close to 0.6 V ?
I also would like to switch on the amp with a relay. I bought two, just for testing. One is a solid state relay and the second one a conventional relay, both rated 25A. I install them on the live.
Is it normal that I can measure AC also on the coil ?
Any "sonic" reason why I should not have a relay on the main ?
Thanks,
Davide
I have my 4 f5 channels done and I have my power supply sorted...
But I guess I've run into the dreaded heat-sink decision roadblock. I had some more creative ideas that ultimately wouldn't have worked out so I've turned to the more traditional solutions. But expense is an issue and from all my research Heatsink USA seems to be the most cost effective for the state-side crowd. I haven't had much luck finding anything on ebay and I don't want the finished product to look like it fell out of the back of a truck...
So do you guys think an 8" slab of this will be sufficient per board?
10.080 inch profile custom cut aluminum heatsink
It looked like it to me but I understand that simply buying a longer piece doesn't necessarily mean it will move enough heat so I wanted to run it by the experts.
But I guess I've run into the dreaded heat-sink decision roadblock. I had some more creative ideas that ultimately wouldn't have worked out so I've turned to the more traditional solutions. But expense is an issue and from all my research Heatsink USA seems to be the most cost effective for the state-side crowd. I haven't had much luck finding anything on ebay and I don't want the finished product to look like it fell out of the back of a truck...
So do you guys think an 8" slab of this will be sufficient per board?
10.080 inch profile custom cut aluminum heatsink
It looked like it to me but I understand that simply buying a longer piece doesn't necessarily mean it will move enough heat so I wanted to run it by the experts.
The par metals extrusions will not dissipate the heat that Heatsink USA sinks can handle. Look at the height of the HS USA vs PM. You dont want the fins ultra close together either, thats for forced air cooling. To tight together and they wont let the air flow up the fins efficiently. Between the two I would go HS USA in a heartbeat.
Uriah
Uriah
Doesnt really matter since the air moving over the mosfet is taking away nearly none of the heat. Facing down your other components would get warmer like you said but probably the only one to worry about would be the thermistor and I assume you are going to put it on or nearly directly on the mosfet itself. The sink is taking it away and the air moving over the sink is what really matters. Place them about one third of the way up from the bottom of the sink.
I forget which site its on, but on one of the major heatsink manufacturers sites there is a java or flash based system that will simulate dissipation for you. So you find a similar profile (it has to be very very similar) and then tell it how big your mosfet is and how much heat it will generate then the program shows you how hot it will be at different points on the sink.
I cant find that tool anymore on either Aavid Thermalloy or on Wakefield's sites. However, if you go to Aavid's site you can choose an extrusion and then it will show you the average dissipation at different air velocities (choose natural convection) and also at different ambient temps.
I think this link will work to show you what I mean
Extrusion Profiles
For good information on heatsinks and frankly to answer just about any question about them in this application you can go here for a lengthy but direct discussion
ESP - Heatsink design and transistor mounting
Pay special attention to how to mount the transistors.
Uriah
I forget which site its on, but on one of the major heatsink manufacturers sites there is a java or flash based system that will simulate dissipation for you. So you find a similar profile (it has to be very very similar) and then tell it how big your mosfet is and how much heat it will generate then the program shows you how hot it will be at different points on the sink.
I cant find that tool anymore on either Aavid Thermalloy or on Wakefield's sites. However, if you go to Aavid's site you can choose an extrusion and then it will show you the average dissipation at different air velocities (choose natural convection) and also at different ambient temps.
I think this link will work to show you what I mean
Extrusion Profiles
For good information on heatsinks and frankly to answer just about any question about them in this application you can go here for a lengthy but direct discussion
ESP - Heatsink design and transistor mounting
Pay special attention to how to mount the transistors.
Uriah
That Rod Elliot site is just awesome. The heatsink article, reading it now, is fantastic. For instance check this out.. Bet you didnt think painting your sink with black oil paint would dissipate more heat than black anodized and consider that your heatsink from both places you mentioned, par metals and HS USA, will NOT be anodized.
Table 3 - Emissivity of Various Surface Treatments
Surface Emissivity
Polished aluminium 0.05
Polished copper 0.07
Rolled sheet steel 0.66
Oxidised copper 0.70
Black anodised aluminium 0.70 - 0.90
Black air-drying enamel 0.85 - 0.91
Dark varnish 0.89 - 0.93
Black oil paint 0.92 - 0.96
He also has a heatsink calculator http://sound.westhost.com//heatsink.zip
Table 3 - Emissivity of Various Surface Treatments
Surface Emissivity
Polished aluminium 0.05
Polished copper 0.07
Rolled sheet steel 0.66
Oxidised copper 0.70
Black anodised aluminium 0.70 - 0.90
Black air-drying enamel 0.85 - 0.91
Dark varnish 0.89 - 0.93
Black oil paint 0.92 - 0.96
He also has a heatsink calculator http://sound.westhost.com//heatsink.zip
Last edited: