June 11, 20206 yr 8 hours ago, thuytn said: How hot does the heatsink get? Seems to me it's like 220-250mm long? That power hole seems to be meant for the 12mm Bulgin-like switch... I have about 160mA bias atm and the heatsink is about 34 Celsius. Transistors are about 10 C higher. I'll push it a bit closer to 180-200 once it is all done to see if it makes any difference.
June 11, 20206 yr 3 hours ago, MASantos said: I have about 160mA bias atm and the heatsink is about 34 Celsius. Transistors are about 10 C higher. I'll push it a bit closer to 180-200 once it is all done to see if it makes any difference. Thank you. Seems much cooler than I thought.
June 11, 20206 yr Author final version of SS/ZF board with jumpers to go from one to the other. 5 db of gain unbal to unbal 11db of gain ubal to bal 17db of gain bal to bal thinking of a relay board to sit on top with front panel SS/ZF switch. would be a unique feature cfp3rssfixedsssw - CADCAM.ZIP
June 12, 20206 yr 51 minutes ago, MASantos said: What is the SS/ZF thing? Super-Symmetry and Zero-Feedback. The input section of this board can do both topologies.
June 12, 20206 yr Thanks! I've listened to the CFA and dynalo mini for about 2 hours this afternoon trying to compare them. I liked the cfa better I believe but will keep listening. They are both great amplifiers but I can only keep one.
June 13, 20206 yr Hi Kevin, the 4 transistors around THAT340 are in different orientation, the THAT340 pnp C is to the following Tr E, while the npn is to C. which one is correct? All ON brand MJE15030, 15031 that I used in the power output had oscillation, some Motorola were stable. Is the fT30 too high?
June 13, 20206 yr Author the 2 n's need to be flipped. this error got there when changing from the 2s transistors to the bc transistors. edit: updated schematic cfp3rssfixedsssw - CADCAM.ZIP cfa3productionss.PDF Edited June 13, 20206 yr by kevin gilmore
June 13, 20206 yr Done! Neutrik 4 pin xlr mounted from behind with blind taps. all in all went well considering it was done all with a hand drill. No drill press around 😔 The on/off indicator led was mounted underneath the case, makes for a nice effect. Quite happy with the result! Now the quest for a better streamer starts!
June 13, 20206 yr Author 15lyl... if you are running ss and don't have the latest board, the feedback might not be correct. Check your board with the schematic posted. i don't know of any way to make ZF oscillate
June 14, 20206 yr I etched the board myself, am running the ZF. I can only stop the oscillation by connecting a 500P to the B, C of the output transistors, smaller than 500P can’t stop the oscillation. I may try to run the SS later to see the negative feedback will help the oscillation.
June 14, 20206 yr Well, I'm no EE, but generally feedback is what lets amps oscillate. Hence why KG said he doesn't know how to make a ZF amp do it. What makes you think that the amp is oscillating?
June 14, 20206 yr Author It is actually possible for something with no voltage gain and no feedback to oscillate, but it's difficult to do so. a lh0032 with the input tied to low impedance ground will oscillate.
July 1, 20205 yr There are still missing parts but so far so good, transistors will be all mounted directly on the heatsinks. This is gonna be a dual chassis setup with custom cut heatsinks. I have some balanced headphone protector boards, not sure it’s worth adding the extra wirings... still waiting for Kevin’s ZF/SS switch board anyway.
July 1, 20205 yr I'm adding a protector to mine, will run it from one of the GRLV's. The shutdown on mine sends a 2VDC transient out, so I'd better be safe, never know when power can go out and you won't be able to unplug the phones.
July 2, 20205 yr I also had the protector powered off the GRLV in my Dynahi. It worked fine but the Dynahi itself is very stable. I'm not sure with stock parts on the silkscreen, what's the dc offset trigger value? I know amb's e12 will do at 80mV.
July 3, 20205 yr I always wonder what makes the CFA2 sound so good and makes me choose it over the dynahi every time. Does the magic lie in the current feedback topology or the triple darlington output stage? just an idea - what happens if we parallel two or more sets of the triple darlington output for each CFA2 board? Would it be the more the merrier?
July 13, 20205 yr While waiting for the delivery of the heatsinks, I built 2 short DC umbilical cables using 18awg wires. I experimented with new cable sleeve and the result is just great, very soft and flexible.
July 14, 20205 yr 7 minutes ago, Mirko said: @RudeWolf Nice build! What transistors are using instead of that340? A pair of 2sj109/2sk389. Sounds better that way and the DC is way more stable. I'll add that even with the THAT340 it sounded superb! Edited July 14, 20205 yr by RudeWolf
July 19, 20205 yr Hi Kevin, Thank you for your message about the bias of the cfa3 amp, I am running the ZF, by disconnecting the 4 600p, I could adjust the bias smoothly to 200mA, no squeak or oscillation any more. What are those 600P for?
July 22, 20205 yr On 7/15/2020 at 4:46 AM, RudeWolf said: A pair of 2sj109/2sk389. Sounds better that way and the DC is way more stable. I'll add that even with the THAT340 it sounded superb! We cannot get these 2SK109 and 2SK389 anymore, they are obsolete for a long time, all of them are fake from AliExpress.
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now