Underside of the dynalo mini done, just need to clean the flux residue when it's all done. 

Those 0805 resistors are starting to be  the smaller I can solder without magnifying glasses.

I just noticed looking at the picture that the tantalum capacitors don't have the positive to the same side of the board. I must hsve confused this somehow and will take care of it in the morning.

I think IIRC that stripe (+?) o the tantalum goes towards the back of the board.
Also, I wouldn’t populate the 0 ohm jumpers until you have tested and adjusted the amp.


Sent from my iPhone using Tapatalk

56 minutes ago, Pars said:

I think IIRC that stripe (+?) o the tantalum goes towards the back of the board.
Also, I wouldn’t populate the 0 ohm jumpers until you have tested and adjusted the amp.


Sent from my iPhone using Tapatalk

They should all point to the same side indeed, but I for some reason I can't understand I soldered them this way... ?

Regarding the 0ohm jumpers, I didn't understand I should only install them after testing the amp, I'll remove them tomorrow as well then.

It seems I have forgotten to order the MMBTA56 SOT 23 transistors. Does anyone have 6 of these by any chance that they would be willing to sell me?

Ordering from mouser will cost me an arm and a leg in shipping and customs for 1 dollar worth of parts.

@MASantos

i need to order some parts for myself soon. I can add the MMBTA56 transistors to my order

thanks, that would be great. I'll send you a pm later today.

@cspirou 

Arrow has free international shipping, so I just ordered from them. Thanks for you help anyway!

Guys, a question for those who used AMB's o78/79 regulators.

Do I need to isolate the PCB from the heatsinks? 

Yes


Sent from my iPhone using Tapatalk

Hi All,

I have my first of two Dynalo minis populated. I fell into the same trap as Congo with the orientation of the 680 ohm resistors, but once fixed it came up without issue. I had already checked and tweaked the on board regs.

My question is about the biasing. I started with the default recommended 422ohm and 255ohm. Then I tried adjusting the combinations as others have before, and I am summarizing what I found into a table. However, I seem to have the opposite effect to Pars, which is confusing me. I have the Liteon LED's, and +/-20V supplies.

With R1,R3 = 422ohm, R14,R29 = 255ohm: 13mA

With R1,R3 = 402ohm, R14,R29 = 255ohm: 16mA

With R1,R3 = 422ohm, R14,R29 = 240ohm: 11mA

With R1,R3 = 402ohm, R14,R29 = 240ohm: 13mA

Can someone help explain this?

Attached document is my collection of bring-up notes and some copy-pastes from the forums.

https://docs.google.com/document/d/1iJ_jL2ojQpJ2f_3crzrbiI6Gti7ZWHiXPHGWbkdjHDA/edit?usp=sharing

I think I let the magic smoke out by swapping into JFET input. Now the amp functions normally with input on + and G input, but not - and G.

Should I worry about anything else than the J109? (another pair of J109 and K389 worked perfectly)

5 hours ago, Flyingsparks said:

Hi All,

I have my first of two Dynalo minis populated. I fell into the same trap as Congo with the orientation of the 680 ohm resistors, but once fixed it came up without issue. I had already checked and tweaked the on board regs.

My question is about the biasing. I started with the default recommended 422ohm and 255ohm. Then I tried adjusting the combinations as others have before, and I am summarizing what I found into a table. However, I seem to have the opposite effect to Pars, which is confusing me. I have the Liteon LED's, and +/-20V supplies.

With R1,R3 = 422ohm, R14,R29 = 255ohm: 13mA

With R1,R3 = 402ohm, R14,R29 = 255ohm: 16mA

With R1,R3 = 422ohm, R14,R29 = 240ohm: 11mA

With R1,R3 = 402ohm, R14,R29 = 240ohm: 13mA

Can someone help explain this?

Attached document is my collection of bring-up notes and some copy-pastes from the forums.

https://docs.google.com/document/d/1iJ_jL2ojQpJ2f_3crzrbiI6Gti7ZWHiXPHGWbkdjHDA/edit?usp=sharing

I probably got it backwards as from what I recall, the 402R biased it hotter than the 422R did. IThe 15mA or so came after it was running for at least 30 mins, more like an hour. Putting the heatsinks on should drop it a bit (couldn't measure, but based on overall current draw of the amp).

EDIT: just played around in a spice model, and lowering the LED resistors (R1, R3) biases things hotter.

Edited December 30, 20187 yr by Pars

35 minutes ago, lkong said:

I think I let the magic smoke out by swapping into JFET input. Now the amp functions normally with input on + and G input, but not - and G.

Should I worry about anything else than the J109? (another pair of J109 and K389 worked perfectly)

You should be OK, assuming you are using the right adapters (don't use the SA1349/SC33xx adapters!). There aren't really any voltages present there that would damage the JFETs, unless the Gate is wired incorrectly.

16 hours ago, Pars said:

You should be OK, assuming you are using the right adapters (don't use the SA1349/SC33xx adapters!). There aren't really any voltages present there that would damage the JFETs, unless the Gate is wired incorrectly.

What worries me is what caused it, I simply tested the amp with 1k sine wave. I hope it's just the input jfet.

 

Edited December 30, 20187 yr by lkong

Thanks for the confirmation Pars. Seeing you have the spice sim set up, does the drop of R14,R29 from 255 to 240ohm also result in a drop in Bias? Looking at the circuit and thinking about the behaviour, it makes sense to me. I just got rather confused by past posts.

This started as a wish to collate the results posted of different configurations. For completeness I will probably initially populate the 2nd amp with the Bivar LED's, as I ended up ordering both.

Cheers

Yes, dropping R14/R29 drops the bias level. Sorry for any confusion in past posts.

No problems, thanks for clarifying. I think you had suggested in previous posts to collect the info together, so that is what I'm trying to do.

Now I have to wait for the heatsinks to arrive.

@pars some measurements for my finished second smy dynalo - found a good combination for the bivar leds.

amp gets stable after 30 min (with heatsink mounted), hottest parts are the small transistors at ~81 c, the regulators are ~75 c.

heres the data:

D1 / D2s = Bivar SM0805HCL
R14 / R29 = 226 ohm
R1 / R3 = 402 ohm
330mv / 16.5 ma hot
205mv / 10 ma cold

my first dynalo has R14 / R29 = 240 ohm and was running to hot, could bring the heat down with some heatpipes attached from the
heatsinks to the case (power consumption stable at 25-26 watt). the second one runs at 20-21 watt without heatpipe.

i also change the trafos to the mean well IRM-20-24, same size but 2x21.6w, even a tad cheaper. 

Updating my BOM for the Dynalo Mini.

MMBTA56/06 looks to be discontinued. As a substitute Mouser is recommending MMBTA56L/06L. Is there any reason I shouldn't use this substitute for the Dynalo Mini? The datasheets seems to match on most specs and the graphs differ a little bit

EDIT: Oh god 32 week lead time! Fuck that!

Edited January 23, 20197 yr by cspirou

try 2SC3324/2SA1312

They have both the replacements in stock -200,000 of each or more.


Sent from my iPhone using Tapatalk

21 hours ago, Pars said:

They have both the replacements in stock -200,000 of each or more.


Sent from my iPhone using Tapatalk

My mistake. I just focused on the red text saying "Please note long lead time"

I fired up the mini this morning, no smoke or fireworks which is always great! 

I left it for half an hour to warm up with the case closed and measured the temperature over the transistors with an IR thermometer. I got temperatures between 90-100 celcius, which is very hot.

This seems to  be quite hot, much higher than what other members reported. 

Is this safe or should I look into reducing bias? I don't mind the heat, but don't want to have to replace transistors anytime soon.

I would measure the bias on every 20R resistor and go from there. Note I did say every, and I wasn’t joking. I have had bad solder joints which effectively took a transistor out of the mix; these looked good to the eye, so you need to make sure they are all working correctly.
Normally on a full size Dynalo I like ~17mA bias. On the mini, because of the thermal considerations, 15mA is the highest I would go. 13-14mA would be a good target. Sub 10mA is very audible, and too low in my opinion.


Sent from my iPhone using Tapatalk

Mine is still running at 100-110C on the transistors and I accidentally left it on for 2 hours and 30 minutes the other day.  It is still singing though