It is indeed possible to upload a sketch in the Arduino IDE with an USBTinyISP (Ctrl+Shift+U). I just wanted to point out why that is a bad idea.
However; it is good to find out that it wouldn't involve too much un-soldering to fix.
I had no idea it was a problem until I wondered why my Arduino UNO would no longer accept programming from it's built-in serial interface.
I was eventually able to get good soldering results. It is just something I wouldn't recommend anyone to attempt without at least a temperature controlled iron and a flux pen.
Hopefully I'll have some time soon to finish populating my PCB and test it.
Looks good, I might have to get some when they are ready.
Even though is is probably now too late to suggest a change, it would be nice to have a provision for a RC snubber on the PCB to protect the power relay contacts. There is a large amount of back EMF when opening contacts to a inductive load such as a transformer. Using a snubber will allow the relay contacts to last much longer.
I was alarmed to not see reverse biased diodes for the relay coils for the same reason. Until I realized the relays are being driven by a uln2003. Those have a built-in diode.
I also managed to get my SMT version running, I had to wait for a new TC54 to come in the mail, the first one I installed end up being bad. The input was shorted to ground.
Even though is is probably now too late to suggest a change, it would be nice to have a provision for a RC snubber on the PCB to protect the power relay contacts. There is a large amount of back EMF when opening contacts to a inductive load such as a transformer. Using a snubber will allow the relay contacts to last much longer.
I was alarmed to not see reverse biased diodes for the relay coils for the same reason. Until I realized the relays are being driven by a uln2003. Those have a built-in diode.
I also managed to get my SMT version running, I had to wait for a new TC54 to come in the mail, the first one I installed end up being bad. The input was shorted to ground.
I could add provisions for a snubber, but I don't think it's really necessary. These relays are designed to handle DC switching and are also horsepower rated. The little transformer load is really nothing for them, especially with a softstart removing the big inrush.
Its been a year since I've built my tube amp if you remember. The amps are still on the bench after destroying my shoulder last year. I made some upgrades replacing the IT transformer and replaced the driver tube with a Russian 6E5P in triode mode with CCS. These amps are singing now and perform better than a Cary 805.
I noticed you have a 4 relay version of this board now, have you done any more tube amp protection boards lately?
Have you done any more programming on tube amp supplies? I haven't mounted those other boards you sent me and saving them for another project (really needed the 4 relays).
I have since developed a problem with my Arduino board dropping out some relays. I think its heat related since the UNO is mounted upside down.
I was looking at some pictures of your amp a couple weeks ago. I was wondering if you ever got it finished.
Your 12V regulator might be overheating and dropping out to save itself. I've added power saver circuits to the newer designs to cut down hold in current on the relays to make everything run cooler in the later designs. I was able to decrease the control transformer size doing this. These have come a long way since the version you have.
I've got a couple different versions of tube hybrid protection systems running. The four relay board is actually two outputs with softstarts. I've also got a control board with 5 relays on it, but it's two outputs with softstarts and an auxiliary out.
Do you need four separate outputs? Let me know exactly what you need and I can likely put something together.
I thought it was the regulator as well but just a couple of the relays drop out (unless the regulator starves the current and by luck the two HV relays drop out)
Sequence:
Pushbutton 1-starts Arduino program (normally open button)
LED 1 on- Warmup LED indicator
Relay 1- Filament supplies
60 Second delay
Relay 2- High voltage to 300B, CCS to 6E5P
15 second delay
Relay 3- Softstart to GM70
5 second delay
Relay 4- Softstart bypass relay
LED 1 off, LED 2 On (operation LED)
Have not implemented protection yet, but intent was to start monitoring cathode voltage immediately after relay4, if out of range (110-120v) then power down sequence and turn on LED 3 protection light.
May do the same monitoring of the CCS to 6E5P (set max at 190v)
Sequence:
Pushbutton 1-starts Arduino program (normally open button)
LED 1 on- Warmup LED indicator
Relay 1- Filament supplies
60 Second delay
Relay 2- High voltage to 300B, CCS to 6E5P
15 second delay
Relay 3- Softstart to GM70
5 second delay
Relay 4- Softstart bypass relay
LED 1 off, LED 2 On (operation LED)
Have not implemented protection yet, but intent was to start monitoring cathode voltage immediately after relay4, if out of range (110-120v) then power down sequence and turn on LED 3 protection light.
May do the same monitoring of the CCS to 6E5P (set max at 190v)
I have a smd heat gun and have reflowed video chips in laptops so should be a breeze. I have a decent to work, 4x6 and another 3x6 area
Attached some audio porn
Red glow is led bias on 6E5P
Attachments
Hot air make these really fast and easy to assemble. The amps looks cool! i need to learn some more about tube amp design and build a tube amp some day.
My big board is just a little too large to fit in a 4 x 6 area. I'll see if i can put together a layout that will work for you this weekend.
What are you using for a softstart resistor? What's the current draw of each of your supplies?
My big board is just a little too large to fit in a 4 x 6 area. I'll see if i can put together a layout that will work for you this weekend.
What are you using for a softstart resistor? What's the current draw of each of your supplies?
When you get ready to learn tube amps, I'll send you a one drive link to all my ltspice files including the executables, libraries and models. I'm still learning every day and just implemented CCS on my driver tube. I have tons of example schematics in spice.
Filament supplies are about 1 amp
2 relay is about 1/2 amp
HV relay is about 1-1/2 amps
I put spacers on my arduino board tonight and it took a lot longer (2hrs) to shutdown (ambient temp is also a little cooler).
Must be heat on regulator but puzzled why it doesn't drop out all the relays. Must be coincidence why just the last two relays drop. Unless the arduino has a built in shut down that load sheds starting with last event?
Ever try to glue a heatsink to the regulator? Or mount a TO220 regulator?
No, I still have the UNO boards installed.
I'm using the boards from last year on my 300b amp since it has one supply plus filament.
I just epoxied a small sink on the SMD regulator tonight and will give it a try.
The regulator doesn't pass the thumb heat test (it will burn you) so it really needs to be sinked.
Did you check the legibility of your posted pic?
Are you just trying to waste our time?
This is what we are trying to avoid. No fixed logic.
High quality sensors + microcontroller with well-tested adjustable firmware = combination of reliability and flexibility.
We also use solid-state relays (except the ones for the mains connections 110/220V AC) - mechanical ones are not reliable enough at high currents.
As a result we provide:
1) Flexible soft-start sequence (1 to 4 control relays) - firmware controlled;
2) Protection with the following sensors:
- DC offset (as many channels as required);
- over-current at the output stage (2 channels in basic configuration);
- over-temperature of the heatsinks (I2C sensors, as many as required);
- AC failure (important for high-power amplifiers).
We also provide the option of the power rails cut-off protection for high-end implementations, protecting the output stage and PSU from damage.
We also indicate the reason of shutting-down the amp - many options from a single LED to LCD screen. Next generation of firmware will allow showing the key operation parameters on LCD.
In overall, this is a complex control solution, utilizing redundancy and speed acceleration circuits and firmware features for bullet-proof operation.