mwl168

On my Atmasphere OTL amp, there are two power switches. One powers the first stage tubes (6SN7) fully and only the filament of the output section tubes (6AS7). The second one supplies the B+ to the output section 6AS7. The procedure is to turn on the first switch and wait at least 60 seconds before turning on the second switch. The switches are wired such that the second switch is only effective when the first switch is on. Turning off the first switch cuts off the power to the second switch. Seems very similar to what JimL described about the T1. It was implemented this way to preserve/extend the life of the 6AS7. It's manual but it works very well for me.

What did you not like about the box? I am contempating getting a smaller one for my Carbon build.

that's a great idea!

I have been informed of a few PCB issues from a few GB participants in the last few weeks and want to bring this information to your attention. Please do let me know if you run into similar issues or other issues which you believe are related to the PCB. I like to file a complaint with the manufacture with a summary of the issues and get their response. This manufacture (EMB PCB) was recommended to me by a forum member and this is the first time I’ve used them. I should also note that my order with EMB PCB includes other PCBs other than the KGST ones. Here is a summary of the issues that have been reported to me but they are not specific to the KGST boards: 1. One report of an entire island of solder came off during de-soldering with solder wick 2. Two reports where a few through holes were either too small or over-plated making inserting the parts difficult 3. One report of a couple of spots on the PCB where the copper is visible through the solder-mask maybe the solder-mask didn't take properly 4. One report where one corner of the silkscreen appears smudged 5. One report where 2 through holes were completely plated an need to be drilled open. For issue #2, I tried inserting parts in the same spots on the board I have and did not encounter problem. So this appears to be a QA consistency issue. I should also mention that sorenb who encountered the problem with the solder island coming off has built a working KGST amp with the boards he got from this GB. However, the solder island should not have come off so easily. Again, please let me know if you run into issues or if I have missed issues that you have reported to me. I apologize for these problems and will share with you what response I receive from the manufacture when I file the complaint. Thanks!

The "fat" single PS was released after the GB was started. I believe the only difference between the "fat" and "skinny" single PSU is the dimension of the PCB based on my quick look through gerber file viewer. Kevin will have to confirm. My guess is the fat single PSU was created to facilitate flexible boxing layout and options. I need to ask all those that have expressed interest in the single PSU to edit your response and specify specifically if you want the "fat" or "skinny" single PSU if you have not already done so. I will be getting a spreadsheet listing the participants and options started and posted her at the appropriate time. Thanks!

Sorenb brough the issue to my attention as well. I do plan on using a different fab house for this GB just don't know which one yet. Please do let me know if you run into PCB quality issue with your KGST boards. I will file a complaint with the manufacture.

Thanks Kevin. Should I re-trace the dual power supply board?

No sure if I understand your question correctly. The original Telefunken made two 12AX7 equivalents, ECC83 and ECC803S. ECC803S has a different physical construction than the 12AX7 and ECC83. IIRC, it has an extra getter and looks similar to 6DJ8. They are extremely rare and pricy now. I read somewhere the ECC803S were built for early computer applications and are guranteed for section to section match.

As promised, I started a separate thread for the KGSSHV Carbon and the GoldenReference HV PS group buy: Due to the options for different PS boards, I need to ask those that have already expressed interest in the GB here to read the new thread and post your response again over at the new thread. I am also going to need some help checking the boards. Thanks!

EDIT 8/4: attach updated GB details and remov outdated chart EDIT 7/31: attach updated GB details and remov outdated chart EDIT 7/27: attach GB details and remov outdated chart EDIT 7/23: attach PCB cost quotes and limit first time poster to 4 set of amp boards max. EDIT 7/19: proposing use of HV Carbon V5 board to replace both V2 and V3 boards. See posting # 190. EDIT 7/13: attach snapshot of final order entries and board count. EDIT 7/12: This GB is officially closed as of 7/12/2015. Due to the changes Kevin made to the PS boards, it'll take a few more days for these boards to be checked before I officially request quotes from the fab houses. I hope to have information on the cost of the various boards to share in a week or so. I will post an image of the spreadsheet here for all to review. EDIT 7/9: Please double check and confirm or make any final updates to your record on the spreadsheet by end of day Saturday, July 11. On Sunday, July 12, I will officially close the GB, lock down the counts of the various boards, post a snap shot of the spreadsheet on this thread for your review and proceed to request final quotes from the fab houses. As promised, I am planning a KGSSHV Carbon GB which includes both the amp boards and the PS boards. First I like to gauge the interest for the GB which will also help me estimate the cost of the PCB boards. My plan is to make 2mm thick PCB with 2oz copper. (EDIT 0731: HV PS boards are upgraded to 3oz copper) For the amp board, we'll most likely use the latest board, kgsshvcarbonv2.zip and kgsshvcarbonv3.zip that Kevin posted on June 30 which has the cascode CCS and uses mostly current parts including the PZTA SMD parts. This board measures 5.45 x 4.05 inches according to Kevin. THIS BOARD HAS NOT BEEN CHECKED NOR TESTED AT TIME OF WRITING. For the HV PS board, we have the following options (these HV boards contain built-in LV supply): 1. the GoldenReference dual HV (kgsshvpssicfetdual2new.zip that Kevin posted on June 27, 6 x 6.48 inches). I have checked this board by tracing the PCB visually and Kevin has corrected the issues I spotted and updated the schematic accordingly (see Kevin's post #4802 on the KGSSHV thread for the schematic). I cannot guarantee I did not miss something. THIS BOARD HAS NOT BEEN TESTED. 2. the split/single, "skinny" PS boards (one board for each rail), they are kgsshvpssicfetsinglenewleft.zip and kgsshvpssicfetsinglenewright.zip which were also posted on June 28 (3 x 6.48 inches). THESE SINGLE PS BOARDS HAVE NOT BEEN CHECKED NOR TESTED. 3. the split/single, "fat" PS boards (one board for each rail), they are kgsshvpssicfetsinglenewleftfat.zip and kgsshvpssicfetsinglenewrightfat.zip which were posted on June 28 (4.05 x 5.45 inches). THESE SINGLE PS BOARDS HAVE NOT BEEN CHECKED NOR TESTED. For stand alone Golden Reference LV PS, two options (THESE PS BOARDS HAVE NOT BEEN CHECKED NOR TESTED): 1. the dual LV (goldenreferce3.zip) 2. the single LV (goldenrefernceplus4.zip and goldenreferenceminus4.zip) We'll need to at least check these boards before proceeding to have them fabricated for the GB. I will try to find time to trace them but can really use others' help for double-checking. To gauge the interest and to help get an idea of the volume in order to determine the cost, please respond with the following information (be specific) if you are interested in participating in this GB: 1. How many amp boards do you want. (one channel on each board, you need two amp boards for one complete amp.) 2. Which HV PS (dual, single skinny or single fat) and how many of them do you want. (you need at least one dual board or two single boards for one complete amp.) The HV PS has built in LV and bias sections. The separate LV PS and Bias boards are optional. 3. Please use this link to register your request: https://docs.google.com/spreadsheets/d/1GWDWwLfsIDtwn1HB-x76lS6WyWwLd5lwsORFKiUuuEQ/edit?usp=sharing The chart below is for reference only and not up to date. The spreadsheet contains the up to date information. Please let me know if I missed anything. It's advisable to read the the following posts at the minimal for relevant information to this GB: 16, 34, 42, 51, 68, 80, 89, 92, 113. More to come. Thanks! EDIT 6/28: updated information due to the newly released "fat" single PSU and updated gerber files for PCBs EDIT 7/3: added Golden Reference LV and Bias board options and start using sorenb's shared spreadsheet EDIT 7/6: remove early spreadsheet image Note: For the single PS boards, quantity of 2 indicates the participant will receive 1 left and 1 right boards (or 1 plus and 1 minus for the LV).

I do intend to include PSU in the GB. Not sure if to go with the KGSSHV PS Sic version or the GoldenReference HV version. In any case, maybe I should start a thread for the GB and move these discussions over so we don't clog this thread. Just got back from a business trip. Give me a couple of days and I'll get a GB thread started.

Hi Kevin and Birgir: Is there an off-board KGSSHV Carbon version that is tested, finalized and available to the public? If so and if there is enough interest, I can help run a GB. Thanks!

Congrats Sorenb. I think this is the first report of a cascode CCS KGST build. Sound impressions please!

Did some search on Mouser. SMD film caps seem quite expensive and not that small in size. Maybe use polymer filter capacitors and stick with through-hole film bypass capacitors?

Hi Kevin: Is it possible to make the filter and bypass capacitors also SMD? The finished board would have a really low profile and can be stacked together and still fit into a slim case.

Goodness! That's compact! Wiring must've been fun!

It's a used item and not refundable. Wonder what's the key for? Ignition?

impressive work as always John. Congrats! Have you measured the temperature inside the enclosure yet?

Lamborghini started as a company manufacturing farming machines.

Here is a photo as promised.

UCC is a popular DIY supplier and has been around for some time. The dual-rail power supply is available both as a complete kit for roughly $43 US or PCB only for roughly $4 US. I was not able to find the shipping charge to US. It also appears that the store only accepts cash - the buyer is instructed to send cash to the store via registered mail.

The PCB boards arrived today. I'll post photo tomorrow and start shipping them out. Hi Rodeodave: Sorry, but all the boards I ordered are spoken for a few weeks ago. Best!

Thanks for tips from sorenb, I tinkered with the VAS on my KGSSHV this morning and after some try and error it appears to be working now. I thought I would share my experience here. I could not find Kevin’s original KGSS article any more (what’s up with the HEADWIZE site?) but going by memory the CCS for the VAS was supposed to source about 3 mA. I am using the ixtp01n100d part, with the drain resistor (R12/R19) at 1.2K as marked on the silkscreen, I got 2.36V voltage drop – only about 2mA of current. I paralleled R12/R19 with a 1K resistor and got it down to 546R and about 2.35V voltage drop - about 4.3mA current. This change very slightly raised the output device current from 9.7mA to 9.8mA. But two things caught me by surprise: 1. the amps are drawing more current from the PSU, voltage drop on the 15R/3W current-limiting resistors (R2, R3, R13, R14) went from 3.75V to 4.4V. My, admittedly small, heatsinks are running quite a bit warmer (but not hot). 2. the balance and DC offset are now all over the place. To the point I thought I somehow messed up a working amp (see more below). With the original resistors in the offset chain (R26 = 2K, VR1 = 1K) I could not get the balance and offset to anywhere near 0V. Not only they were all measured well over 200V, it was so bad it almost acted like the amp was not responding to adjustment at all. I had to finally parallel R26 with a 1K resistor and brought the value down to 666R. At that point I was able bring the offset to 0V. Once I was able to get the offset in control the balance adjustment falls in place as well. As it sits playing now, offset VR (VR1) is adjusted to about 235R. I don’t like to describe subjective sound impressions. But I do believe the changes I made to raise the VAS CCS current do improve the sound of the amp.

The Item description says it's not guranteed to work since the poster does not have a headphone to try the amp with.

Got the PSU running correctly now - the zener string on the - rail now measures around -445V as should be. In the process I also changed R5/R6 on the amp from 120R to 100R - might as well take advantage of the 290ma current rating of the transformer ) Cannot really say I hear a difference so far with the higher output current. The heatsinks on both the PSU and amp do run slightly warmer now but not by much. Many thanks to Kerry, Justin and Birgir for your help.