Jump to content

DIY mini T2 Build Thread


mwl168

Recommended Posts

Recovered the GRHV,  I take a look  to  the amplifier. Interestingly, all damage is in the area of Servo/OPA27.

It seems that most of the transistors are fine although I suspect that the IXTP01N100D are damaged (continuity between two of their legs). 

Tomorrow I will begin to dismantle the most damaged channel. 

Link to comment
Share on other sites

HC´s number one rule: have more components in your warehouse than Mouser.

Follow the trail of burned components

Certified dead: CL resistor, r12, c2m1000170d, 2x bc557b, 2n3904. njm7915fa

Changed as a precaution: Q7 10m90s. Q8 & Q9 ksc5026m and 24v diode.  

The rest of the components passed the test.

Total: A few euros and a couple of hours.

 

 

Edited by jose
  • Like 2
Link to comment
Share on other sites

Thanks, good to know what components that are at risk. I’m using a KGSShv style PSU and I’ve a small inventory of FQPF8N80C, 30V Zener and 15R power resistors and so far that has been enough. Also omitted current limiter since it hasn’t protected me. 

One other question, @jose. Do you have four separate 6.3 VAC for the tube heaters?

Edited by JoaMat
Link to comment
Share on other sites

The resistance I have in the current limiter is about 3R. It is interesting what you have commented about removing the CL resistor. Perhaps without this resistor the damage would have been greater ¿?¿?¿?¿?¿?¿?

Currently my plan is to build a BH or BHmini PSU just for testing.

As for the heating filaments, my T2 mini is configured with the E88CC filaments in DC (with a separate transformer and a PSU for this work) and EL34 with the "standard" way;  four tubes with AC with its own transformer. Should I use separate transformers?

Link to comment
Share on other sites

I haven’t removed “current limiter” resistor – here I use 15R/3W – just removed the 2n3904 transistor.

I’m only used to 4 separate heater windings so I can’t tell about running all big tubes as you do. Mini T2 has big tubes heaters tied to roughly -400 V via R53-56 voltage divider, so you actually connect left and right channel’s dividing points with your single filament transformer. My brain is to slow to figure out what, if any, implications that might have. Perhaps someone out there can enlighten us.

But never use same heater windings for big and small tubes. Small tubes don’t like heaters to be tied to -400 V and besides their heaters are tied to ground via 47R resistors.

Link to comment
Share on other sites

I believe all the 6922s might be run off one winding since the heaters are ground referenced and the current draw is much less than the el34s but I have not tried it. As joamat says the el34s cant share a heater winding with the 6922s since the el34 heaters are -400V with respect to ground. I have not tried running both channels el34s from a single winding because a. thats a lot of current draw 6A ignoring switch on inrush and magnetic losses b like joamat I don't know if it would be safe to do so.  c separate windings provides a bit more isolation if/when things go wrong. You might want to message Kevin Gilmore for his advice....

P.S. jose have you signed up to the please add screw terminals to the mini t2 amp boards request yet 👿

 

Edited by jamesmking
Link to comment
Share on other sites

2 hours ago, JoaMat said:

I haven’t removed “current limiter” resistor – here I use 15R/3W – just removed the 2n3904 transistor.

I agree with this. The transistor does not provide protection against shorts, so I remove it as well and use the 15R resistor.  When I feel it necessary, I use an external current regulator. 

Link to comment
Share on other sites

I think I have not expressed myself well:

For the EL34 I use a single trafo with two windings. One winding per channel (2 EL34 per winding)
For the E88CC I use another trafo with its own psu for DC heater. 

Both types of tubes DO NOT share transformer.

Edited by jose
Link to comment
Share on other sites

2 hours ago, jamesmking said:

I'm running + and - 15v, 6922 and el34 from the same transformer all heaters are AC. As long as the insulation on the windings can handle the 400V between el34 and the other windings its fine.

That's similar to the DIY T2 and works just fine.

Link to comment
Share on other sites

On 6/1/2020 at 4:34 PM, jose said:

My idea is to remove all parts of that servo and run the amp without it or using the servo on top. I understand that it is possible.

To find out I removed Q10 (also removed vol618) and replaced it with 1K trimmer (just put it in Q10 position) to control the offset. With balance trimmer (jumper removed) and offset trimmer I’ve no problem to manage the amplifier, but this is just after a day. Maybe I tell more later.

Link to comment
Share on other sites

40 minutes ago, JoaMat said:

To find out I removed Q10 (also removed vol618) and replaced it with 1K trimmer (just put it in Q10 position) to control the offset. With balance trimmer (jumper removed) and offset trimmer I’ve no problem to manage the amplifier, but this is just after a day. Maybe I tell more later.

Thanks for advice Joachim!

I´ll work in that direction.

Link to comment
Share on other sites

3 hours ago, JoaMat said:

Think I like trimmers. Instead of servos – miniature stepper motor driven trimmers controlled by Arduino?

Sounds like a battle between the Mechanical Engineer and the Electrical Engineer inside of all of us.  

  • Like 1
Link to comment
Share on other sites

A detail that I forgot to comment, the slow 1A fuses that I use at the output each HV line of my PSU were blown.

The fuse on the HV + rail protected its side of the GRHV but the one on the side - it wasn't so lucky.

Link to comment
Share on other sites

On 6/5/2020 at 11:47 AM, JoaMat said:

To find out I removed Q10 (also removed vol618) and replaced it with 1K trimmer (just put it in Q10 position) to control the offset. With balance trimmer (jumper removed) and offset trimmer I’ve no problem to manage the amplifier, but this is just after a day. Maybe I tell more later.

A week without offset servo and most of the time jumpers for balance servo have been removed. Today amplifier started out with offsets up to -40V. After half of an hour offsets between 0V and -10V and unbalance 5V and 10V. I think this is acceptable.

  • Like 1
Link to comment
Share on other sites

  • 1 month later...

While working with latest mini T2 (version .23) boards today I discovered that op amps on left channel have wrong (old) values. The op amp on none tube side, close to VOL168A, should be OPA197. Op amp on the other side should be LF356. Silkscreen right channel seems to be correct.

  • Like 1
Link to comment
Share on other sites

Warning!

0.1u/1kV film cap pads on right channel have 80-30 pads meaning hole diameter is 0.030 in or 0.76 mm. This is to tight for most film caps that have lead diameter of 0.8mm. There are some Panasonic, Cornell Dubilier, Nichicon and Vishay caps with 0.6 mm leads and 15 mms lead spacing.  Do not drill the holes for the film caps. The through hole plating is needed as it ties ground planes together and also needed for -460V to be fed to none tube side of the board.

Left board have 90-40 pads – 40th or 1.0 mm. 

This mistake is done by me and the blame is entirely on me. I’m sorry.

Link to comment
Share on other sites

Now completed mini T2 with the new boards version .23. 

Squeezed in film caps with 0.8 mm leads on right channel with 0.76 mm hole diameter, elevated and soldered both sides (see below).
IMG_0324.thumb.jpg.265e198452d3bae9b70e06b09cd70ea2.jpg

Both channels worked right away. Only issues I’ve found with those new boards are mentioned in posts above.

 

Edited by JoaMat
  • Like 6
Link to comment
Share on other sites

  • 4 weeks later...

Here are some distortion figures and harmonic analysis up to 50Khz.

Test setup groove tubes el34 matched, electro harmonix gold pin 6922 matched, 1Khz sine from a 192K 24bit rme soundcard spdif out to a v90 dac to the mini t2 inputs, measured on a keithley 2015thd. distortion graph is the average of 20 measurements. All other readings are not averaged:  Distortion is less than 0.02% up to 500V rms output with second harmonic being the highest and a little bit of third above the noise floor. At just under 500V the 4th harmonic is above the noise floor, above 500V rms clipping starts and at 600V rms distorion is ~2.5%  with the second harmonic at -36db relative to the 1khz signal level. On the right hand top side of each graph under the measurement section is the measurements for the 2nd harmonic in db and mVolts average, min and max over 20 measurements. Noise is the ac voltage not in a harminic bin (i.e. a bin which is an integer multiple of 1Khz). Analysis is with 749 bins going from 20hz to 50khz. A full fft with all bins is included for the 100Vrms output only at the bottom of the post.

I wrote the software myself, its very much a rapid prototype kludgy mess but I hope to make it available to community members with keithley 2015 when the code is not so offensive. At the moment there is no guarantee that it will not open up a dimensional hole and make all your original NOS Japanese transistors disappear.

 

1105861145_t11k100vdistortion.thumb.jpg.8ca41b4689ed7c3533a433ad7213c296.jpg1388388288_t11k200vdistortion.thumb.jpg.ae5ecad3838d9049361d617a278f1dbe.jpg1391844372_t11k300vdistortion.thumb.jpg.e6e026d9f12bd6a276d3d3ca55142a9d.jpg2115333200_t11k400vdistortion.thumb.jpg.484e57581686c36511a5c056c8acd162.jpg844482534_t11k500vdistortion.thumb.jpg.f8853d14d4dbaa84b63ac2f25ff2703f.jpg1836862417_t11k511vdistortion.thumb.jpg.d4c2dba9dc3d0bc70ac31952203a15cc.jpg534397009_t11k540vdistortion.thumb.jpg.c6da3f776c8ffdfc2bc0fda5c90a5d3e.jpg

FFT analysis 1Khz 100Vrms output. 20hz to 50khz

 

999300007_fft1khz100v749bins.thumb.jpg.606eedcde8f52820a47cad5c9f8af526.jpg

 

 

 

Edited by jamesmking
  • Like 5
Link to comment
Share on other sites

  • 3 months later...
On 3/23/2019 at 6:06 PM, JoaMat said:

The only thing is that left channels of all the three amplifiers have a bloody annoying hum during the first minutes after start up. I’ve no idea what causes it.

Lately the left channel humming has increased to become even more bloody annoying. 

But… today I cured it (pure luck - trial and error approach). I have two umbilicals between psu and amplifier. Both carrying 580V bias and then joining at the stax connector. So, I removed left 580V bias wire in amplifier and the hum disappeared. To my ears the amplifier is now dead silent.

  • Like 7
Link to comment
Share on other sites

  • 2 months later...

Finally, my mini T2 is completed. Just want to share my build experience and listening impression for those who are about to build one. Also, a huge shout-out for JoaMat, who conceptualised and designed the amp, provided me with invaluable tips when I was debugging my mini T2 and encouraged me to share my experiences here.

--------------------------------

Pics or it didn't happen:

spacer.png

spacer.png

--------------------------------

Build experience: 

Before this amp, I have built the Carbon and Grounded Grid, as well as quite a few dynamic amps including the SMD Dynalo. Overall, I am comfortable dealing with electrostatic amp voltages, but I have limited experiences working on dense and tiny SMD boards.

I built the SMD Dynalo with a soldering iron, but this method was unsuitable for the Mini T2 due to the density of the components and the fact that I tend to use copious amounts of flux grease when I solder by hand - it would be too tough to clean the flux stuck underneath the tiny components. Therefore, I decided to solder the SMD parts by the solder paste and hot-air gun method.

I used 1 set of mini T2 board, 3 through-hole GRHV boards and 1 delay board (warmup.zip). 

This amp is fast to build so long as you have a decent hot air gun (I got mine from Taobao for around 25 USD), a pair of fine tip tweezers and some patience. I used a stencil to apply solder paste on one side of the mini T2 board first, then populated all SMD components and used the hot air gun to melt the solder paste. It was quite satisfying to see the tiny parts drawn into their rightful place by the magic of surface tension. It took me about 1 hour to complete the SMD soldering for 1 side, so I'd say that this method is really efficient for the number of parts to be populated.

The through-hole components and the tube sockets were populated next, and I used some terminal blocks for the power rail inputs. Just be sure to apply jumper wire for the -15V near the servo opamp. Otherwise, you risk seeing 400V at voltage offset, and frying some input tubes.

The transformers I used were 130W and  87W for the HV and LV respectively. There is also a dedicated floating 12V winding for the delay board, and separate filament supply windings are used for the power tubes and input tubes for each channel. Voltage rails were pretty standard - +-15V, +220V, +400V and -460V. The warm-up time was set to 25 sec.

There is no need to adjust for output current bias as in the Carbon or GG. The balance servo is activated by default, and all there is to adjust is the output voltage offset with respect to ground, and it can be easily adjusted so long as the potentiometers are soldered to the same side as the tube sockets.

The amp runs cool. I used heatsinks that measure 7cm tall, 30cm deep, 5cm wide for each side of the chassis. At steady state, the amp module runs at 39 deg C, PSU runs at 35 deg C. In comparison, my 450V 20mA biased Carbon runs at 48 deg C with 15cm-tall heatsinks.

Here are some errors I encountered here and the directions I took to debug. They are based on spice simulation by JoaMat. Hopefully, my tips can help builders who encounter similar issues as I did.

1. On power-up, offset voltage rises to 400V immediately after the HV kicked in. Check the jumper connection for the opamp -15v power supply. They are very short and can be easily missed.  

2. 50K resistor explodes (either R11 or R12). It is likely due to an unconnected input tube or dead tube. Check input tubes and Q1A/B, Q4/5, Q11/12, Q3A/B of the same side. Also check tube sockets, because even the expensive ones can have loose connection on NOS tubes.

3. Q1A/B burns. This might not be easy to spot. The sign can be as subtle as a discoloration/fading of the silkscreen on the component surface, or a subtle exudation of flux around the collector pin of the component. This is likely due to a shorted input tube. 

------------------------------

Sound impressions

Take this section with a pinch of salt, for every DIY build is one-of-a-kind, and mine is no exception. For my mini T2, I used Duelund tinned copper wire as input cable, EIZZ stepped attenuator, STAX SPC earspeaker wires as output wires and a 5-pin socket ripped from a STAX extension cable.

The headphone I used was the ES1a by ES Labs. I wish I had the Stax flagships to do this impression and comparison. The amp is fed by a AK4497 DAC, straight out of DAC and a coupling capacitor, bypassing any low-pass filters.

For the following impression, I used decent tubes - Mullard xf2 EL34 as output, Brimar ECC88 as V1 and Sovtek 6922 as V2. Tubes do affect the sound significantly, and I will talk about that later in the discussion. My 450V 20mA bias Carbon with GRHV&LV will be a point of reference, as I am sure that this is a sound that many here are familiar with. It has the same volume pot as the mini T2 but SPC input wires and copper output wires.

The mini T2 is full sounding with a natural tone. It has got a certain “WOW” factor that makes my carbon sound unimaginative in comparison.

First off, the tonality of the mini T2 is warm and mildly bright such that the sound is airy but non-fatiguing, no matter how loud I cranked the volume up. The bass extends deep with a sufficient rumble where the track calls for it. The airy top end gives the amp a wide soundstage, that is paired with sufficient depth. Instrument and vocal placement are accurate and appropriate, without excessive forwardness or recess. Despite the soundstage being big, it does not sound hollow, because the music is full of details and creates a sense of well-layered space that is almost holographic.

I have tested the amp briefly with other tubes and found that input and output tubes affect the sound most significantly. If input tubes have slightly mismatched sections, the balance would be too great to be zero-ed with potentiometer and balance servo. The small tubes I used are unassuming. Had I upgraded to better tubes, I guess the amp would sound better.

Compared to the Carbon, the mini T2 sounds more organic, more layered and the mids are considerably warmer. It is also way more forgiving on bad recordings. The timbre sounds truer to life, soundstage is slightly larger, and feels like the music has more room to breathe. However, bass extension is not as deep as on the Carbon, and Carbon generally feels more analytical and snappier.

To conclude, if there’s one word to describe this amp, it would be “fun”. It’s fun to build and definitely fun to listen to and roll tubes. I consider it as a step up from the Carbon when I use some decent tubes. Otherwise, this amp with standard new-production tubes would be at least on par as the Carbon.

Edited by dingding123
  • Like 14
Link to comment
Share on other sites

  • 1 month later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.