I will definitely have to take a serious look at local companies, even with the increased price. Just not worth the headache. No fault of the manufacturer, but that is life.
I'd say wait a bit more. The package is very unlikely to be lost, probably just waiting in a BIIIIG line to get processed!
Boards are in. I will build the first set up tonight to confirm proper operation and then begin the shippping process. I hope to ship everything by friday, fingers crossed.
Sorry for delay. Due to ice storm, had no power through Tuesday and just got internet back at the house last night. All is soldered. Will put on wood blank, wire up transformers, and give it a go today. Ill send out board Monday and kits will have to wait as i somehow managed to forget one of the trimmer values and the small diodes on the PSU regs. The parts should be in by Tuesday, so all lkits will ship by Wednesday.
Attachments
Last edited:
Ok.
Boards are finished and making music. There are 3 issues specific to board layout:
1. The silkscreen is wrong for the Mur 860 diodes. They need to be reversed.
2. The intput ground pin is routed through the 100k shorting resistor. This means that the two input holes have to be shorted together and the ground from the signal run to the ground point of the psu input spot. Not a big deal bu we will discuss.
3. The 1K power resistor was layed out wrong, attaching before the 15R rail resistor, not after. THis can be easily resolved by simply by soldering it to the cap leg that is nearby. None of the issues above are serious and even significant. The alterations are basically unoticable if you did not know to looke for them. That being said, if you want new boards with these fixes, then I will make more.
There are two operational issues I have found.
First one is FE offset. For me build, I had a difference in rail voltage of about 1V. FOr whatever reason, this leads to an offset of -6V on FE output. This can be trimmed out by placing a 3K resistor across the 1K 2W resistor on the positive rail. I am investigating whether we can trim the rail difference on the PSU board, leaving the FE components unaffected, as this would be my preferred option. I am consulting Juma's opinion on this. Any other are welcome.
The second is the buffer offset. With the current BF862, I can only get offset to 200mV. This is a fairly easy fix and is resolved by being peculiar about Jfet Idss choice or by lowering the the current setting resistor on the bottom current source.
Once again, I am investigating both options and will present what I believe to be the easiest solution. If the Idss range of the BF 862's is favorable, it may be of no concern, otherwise it will simply mean a resistor change.
All that aside, I was pleased with the sound. It did not really change much, other than adding gain, and IMO, that is all that I want it to do. I want my poweramps to determine my sound, not my preamp. This preamp fits the bill. I will put it into my nicer system tonight and give it more listening time. FE is opp0105 and amps are M2 clone, J2 clone, and Funny^6. I will also work out possible work arounds for above issues. In truth, the operational issues would be present in any implementation, so they are just part of the circuit.
Boards are finished and making music. There are 3 issues specific to board layout:
1. The silkscreen is wrong for the Mur 860 diodes. They need to be reversed.
2. The intput ground pin is routed through the 100k shorting resistor. This means that the two input holes have to be shorted together and the ground from the signal run to the ground point of the psu input spot. Not a big deal bu we will discuss.
3. The 1K power resistor was layed out wrong, attaching before the 15R rail resistor, not after. THis can be easily resolved by simply by soldering it to the cap leg that is nearby. None of the issues above are serious and even significant. The alterations are basically unoticable if you did not know to looke for them. That being said, if you want new boards with these fixes, then I will make more.
There are two operational issues I have found.
First one is FE offset. For me build, I had a difference in rail voltage of about 1V. FOr whatever reason, this leads to an offset of -6V on FE output. This can be trimmed out by placing a 3K resistor across the 1K 2W resistor on the positive rail. I am investigating whether we can trim the rail difference on the PSU board, leaving the FE components unaffected, as this would be my preferred option. I am consulting Juma's opinion on this. Any other are welcome.
The second is the buffer offset. With the current BF862, I can only get offset to 200mV. This is a fairly easy fix and is resolved by being peculiar about Jfet Idss choice or by lowering the the current setting resistor on the bottom current source.
Once again, I am investigating both options and will present what I believe to be the easiest solution. If the Idss range of the BF 862's is favorable, it may be of no concern, otherwise it will simply mean a resistor change.
All that aside, I was pleased with the sound. It did not really change much, other than adding gain, and IMO, that is all that I want it to do. I want my poweramps to determine my sound, not my preamp. This preamp fits the bill. I will put it into my nicer system tonight and give it more listening time. FE is opp0105 and amps are M2 clone, J2 clone, and Funny^6. I will also work out possible work arounds for above issues. In truth, the operational issues would be present in any implementation, so they are just part of the circuit.
Last edited:
The uncrippled F4 can tolerate an input offset of 200mV since it is designed to accept +/-20V input range. Of course, the crippled F4 doesn't care since it is totally capacitor coupled.
Ok.
With help from Juma and Ilquam, I tracked down the problem. The BC550 bipolars on the lower half of the FE was inverted. Switched them out, removed the added resistors, and all is looking good. Offset is about 200mV and this can be trimmed if necessary, but the cap will take care of it before the buffer. In truth, even without the cap, it does not present a problem to the buffer.
The offset on the buffer was the result of the same error on the bc550 as well as BF862's with slightly lower than normal Idss. If their range is consistent, it may mean changing the value of R21 to put the pot in the range to needed to zero offset. I will post pictures of finished boards showing the changes needed and will provide a writeup for all who ordered them. Feel free to ask any questions here, once you start building.
With help from Juma and Ilquam, I tracked down the problem. The BC550 bipolars on the lower half of the FE was inverted. Switched them out, removed the added resistors, and all is looking good. Offset is about 200mV and this can be trimmed if necessary, but the cap will take care of it before the buffer. In truth, even without the cap, it does not present a problem to the buffer.
The offset on the buffer was the result of the same error on the bc550 as well as BF862's with slightly lower than normal Idss. If their range is consistent, it may mean changing the value of R21 to put the pot in the range to needed to zero offset. I will post pictures of finished boards showing the changes needed and will provide a writeup for all who ordered them. Feel free to ask any questions here, once you start building.
Attachments
Good news Buzz. Backward orientation of transistors certainly explains the problems you observed. At some point, post photos of your stuffed boards.
Curly+F4 vs. Impasse+F4 A/B testing
This post reports A/B testing of the Impasse+F4 vs. Curly+F4. The Curly gain stage is a variant of the LSK Pre without an output buffer, intended to drive an uncrippled F4.
I recently finished my A/B preamp+amp testing setup. It consists of two DPDT 24V DC relays to switch the speakers between two amplifier outputs, and a pair of trimmer pot attenuators for the inputs to one of the preamps (the Impasse in this case). The relays are controlled from the listening position with SPST switch at the end of a 16 foot telephone extension cable. The trimmers were carefully adjusted for identical amp outputs at 1kHz. With the gains balanced, flipping the switch is virtually undetectable.
I have listened for many hours to a variety of music and cannot repeatedly detect any differences between the Impasse+F4 and the Curly+F4.
I would like to do similar tests using my STAX Lambda Pro headphones without the F4 amps to better isolate differences between the two voltage gain stages. This could be done by using a stereo RCA switch box to select between the two front ends, and attenuating the output to the STAX by about 10:1.
This post reports A/B testing of the Impasse+F4 vs. Curly+F4. The Curly gain stage is a variant of the LSK Pre without an output buffer, intended to drive an uncrippled F4.
I recently finished my A/B preamp+amp testing setup. It consists of two DPDT 24V DC relays to switch the speakers between two amplifier outputs, and a pair of trimmer pot attenuators for the inputs to one of the preamps (the Impasse in this case). The relays are controlled from the listening position with SPST switch at the end of a 16 foot telephone extension cable. The trimmers were carefully adjusted for identical amp outputs at 1kHz. With the gains balanced, flipping the switch is virtually undetectable.
I have listened for many hours to a variety of music and cannot repeatedly detect any differences between the Impasse+F4 and the Curly+F4.
I would like to do similar tests using my STAX Lambda Pro headphones without the F4 amps to better isolate differences between the two voltage gain stages. This could be done by using a stereo RCA switch box to select between the two front ends, and attenuating the output to the STAX by about 10:1.
While I'm interested in hearing the results of such a test, does that test not actually just test which VAS is a better Headphone amp? I mean, there are different impedances to deal with than driving the F4.... Or is the input impedance of the Stax unit the same as the F4?
I would argue that this peramp is largely transparent. I can easily see Sy's preamp being the same, as it is well engineered. For me, this is exactly how a preamp should be, as i want my power amps to define my sound. If you are looking for a FE for an F4 then i still say it is a good bet. If you want to manuipulate the sound of the power amp, look at the leaving class A artilce. It suggest how one might influence the sound on an output stage like the F4. BTW, I have a pot across the source resistors on my boards. This should allow some level of tuning in terms of final sound. I had to order new pots and smaller resistors for the buffer as the idss range of my bf 862's was on the low side. This will allow for greater adjustabilty in seroing the output of the buffer. I ave also ordered some terminal blocks that will offer some flexibility with wiring in general. Sorry for the delay. I tend to be slow as most of you know. Just trying to be thorough.
I do not suggest using either gain stage for a headphone amp. The only purpose for suggesting the use of headphones was to try to detect sonic differences between the Impasse and Curly gain stages without the influences due to the F4s, the speakers, and the listening room.
lqhuam,
I was just a bit concerned how each VAS will interact with the headphones, as that interaction may influence the outcome. The concern is that if there is a significant difference with the headphones, will that translate to an implementation with the F4?
Perhaps driving the headphones with the F4 in place could be another test?
I was just a bit concerned how each VAS will interact with the headphones, as that interaction may influence the outcome. The concern is that if there is a significant difference with the headphones, will that translate to an implementation with the F4?
Perhaps driving the headphones with the F4 in place could be another test?
- Status
- Not open for further replies.