FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator

[cybermaus]

Yes, still working, even now. (thank god. or invisible pink unicorn. whatever)

When I say I never disconnected the loom from the BP, I mean exactly that. I was of course disconnected from the AWG
But as I had made custom 8 and 4 pin male headers, reconnecting is pretty certain. Even more so now with the vero

[2N3055]

@cybermouse,
Avoid ScanaQuad.. ScanaStudio is hipstery crap (they threw away old software that actually worked, and went to make a new software from the scratch. After almost 3 years they are at roughly 25 % of done)
I was early adopter because it held promise.. If they ever manage to make software to be full featured, maybe then...
4 Ch is good only for serial protocols.. It doesn't support segmented capture and compression..  So very short capture...

DSLogic is much better buy for the money...

Regards,

Sinisa

[DaveR]

Hi soundtec,

I ran mine for an hour with both channels feeding a 15v, 1Mhz sine wave into 50 ohms, lid 90% on, and the temperatures I got were - Cyclone 46C, op amp heat sink 50.5C.  Half an hour later the Cyclone was still 46C and the heat sink was 52.8C, so your FPGA cooler looks like a "belt, braces, old tie, and a piece of string" job .  Nothing wrong with that, as it hasn't cost anything.  I also gave them both the finger tip test, and the Cyclone didn't even feel warm, while the output heat sink was only moderately warm, despite the measured temperature.  Presumably the heat flow is so small that a finger tip immediately absorbs what's available and cools the surface of the chip / heat sink to body temperature, so that there's no apparent difference after the first second or two?  (But now we digress into thermodynamics ...)

Your results of the other output tests will be interesting, especially as mine will be used mostly for radio testing and calibration, when I'll only be using a 0.5v signal fed though an 80dB attenuator - so hardly stressing the op amps at all.

If you want better airflow in the case I think it will have to be either a 30mm fan on the back panel, sucking out, or some holes near the front edge of the lid to allow natural venting (maybe coupled with opening up some of the baffles on the sides, which could be done invisibly).  I still might drill a couple of holes over the LM317 on the 5v module, as it's the only component which really generates any amount of heat in mine, but even that only makes a localised warm spot on the lid and is probably nothing to fret over.

[DaveR]

cybermaus - how many ST-Link dongles have you tried?  I have very little faith in these cheap Chinese affairs as I've now bought six of the things: one died quickly, and only one of the other five will actually program the blue pills, so they're obviously not to be relied upon.

[soundtec]

Maybe the bluepill board is defective Cybermaus.

So the Fy6800 has appeared ,it does seem like Feeltech were keeping an eye on our suggestions all along ,new rubberised Keypad,proper safety earth and it seems , user upgradeable firmware.
I dived in on the competition entry as well ,
Definately worth your while getting in too ,Cyber, DC1,Fremen you guys have done a lot of the leg work on this project.

Thanks for that Dave, interesting result ,the 5 degrees I guess doesnt make much odds either way ,but it wont do any harm either .
Valve music and sound electronics is my game so fans are a big no no for me ,I might consider slotting the recesses on the top panel alright ,with care it should be possible to do it neatly enough . Probably worth your while getting in on the competition too Dave ,very little to loose in any case ,I did add a proviso to the end of my competition entry that the info contained herein wasnt to be given out to third parties for marketing purposes ,finally the world is waking up to data protection and privacy online .Thats interesting radio is your game ,I junked an old microwave oven lately and was thinking about the big transformers contained in them ,and if it would be possible to re-use them as heater and HT transformers for a valve powered transmitter ,the HT would come out at around 3kv from a regular microwave transformer and its possible to rewind them for Lt heater usage quite easily ,an old pc case would work fine as an enclosure ,kinda like a diy preppers valve transmitter made from easily available garbage pile stuff ,just on the off chance the entire global communications network and the internet goes tits up ,

[cybermaus]

I have 3. One China, and 2 more integrated into official distributor bought real ST brand  STM32F0 discovery boards.

As stated, both Bluepills behave the same, both can be programmed to a simple "blink" program and work, both can be programmed to fremen's and then read back to compare without issue. One Bluepill was brand new out of silver anti-static wrapper. So I have very little doubt about the BluePill or STLink themselves.
Nevertheless, I tried once more using the STM32F0 integrated STLink, just to read back. Program verified without problem, was thus properly programmed with the China STLink, and anyway after reprogramming I had same result.

Well, almost same: When connecting both 4 and 8 pin result is the same.
But doing this, I discovered that if I only connect power (4 pin header) and not the 8 pin header, the LED does turn off, it does get past initialization.
Added my pull-up resistor again, but still not work. Something is unstable on that SPI bus when connected.

It may be because I have slightly older hardware. I need to properly spy that bus.
I'd go for the DSLogic, but that comes from china, I do not want to wait 3 weeks. I'll look for something closer by.

[Marco]

I'd go for the DSLogic, but that comes from china, I do not want to wait 3 weeks. I'll look for something closer by.

On ebay there's two Chinese sellers with German warehousing and one seller in the UK.

[cybermaus]

Sellers on eBay and AliBaba that claim to be in UK or Germany are lying.
I know people do actually exist in UK and Germany, but only a few million, and there is like a few billion people in China, so statistically above statement is true.

Also, I noticed the eBay sellers do not include the special coax-probes, but just the straight lead probes.


Edit: More seriously, this one does seem to be in Germany, judging by his claim it will be here Monday and also his offer to drive by and pick it up yourself.  Probably be a "clone" though, judging by the non-coax probes. But I am traveling for work next week, so no need to order that now, I't arrive exactly as I leave.

Edit2: Even found a "Dutch" seller. But: 15 days, so drop-shipper. In fact it says so in the header as if its a good thing: "delivery straight from manufacturer". Why would I pay you €106 instead of $70 if you have no stock and no handling to add value?

[DC1MC]

I have 3. One China, and 2 more integrated into official distributor bought real ST brand  STM32F0 discovery boards.

As stated, both Bluepills behave the same, both can be programmed to a simple "blink" program and work, both can be programmed to fremen's and then read back to compare without issue. One Bluepill was brand new out of silver anti-static wrapper. So I have very little doubt about the BluePill or STLink themselves.
Nevertheless, I tried once more using the STM32F0 integrated STLink, just to read back. Program verified without problem, was thus properly programmed with the China STLink, and anyway after reprogramming I had same result.

Well, almost same: When connecting both 4 and 8 pin result is the same.
But doing this, I discovered that if I only connect power (4 pin header) and not the 8 pin header, the LED does turn off, it does get past initialization.
Added my pull-up resistor again, but still not work. Something is unstable on that SPI bus when connected.

It may be because I have slightly older hardware. I need to properly spy that bus.
I'd go for the DSLogic, but that comes from china, I do not want to wait 3 weeks. I'll look for something closer by.

I wanted to modify a Zeroplus LAP-C 16032, like in the famous thread on EEVBlog,  had to use that useless STTC-5xx serias tip for something, the seller wants 70EUR for it:
https://www.ebay-kleinanzeigen.de/s-anzeige/zeroplus-lap-c-16032-16-kanal-logic-analyzer-wie-neu/836656943-168-7434

Too bad he didn't accept 50, but maybe you want to give it a try.

 Cheers,
 DC1MC

[fremen67]

I am in dire need of a new/better LA

Trying to figure out why it is not working anymore. It is weird, because the BP was never disconnected from the loom, so there is not much chance of me making a mistake with it.
Tried a new BP, and soldered it to a veroboard for better signal and connection quality, and same result. Both BP have the red LED stay on, and both BP nicely run the standard "blink" program from STM32

Using my LA at max 20MHz I am not able to decode what is happening, because the SPI is also at 20Mhz. Ergo, I need a LA of at least 40MHz, better 80MHz
Using my Scope, I was able to see that the SPI signals themselves are pretty clean and well defined, but with the startup delay and limited scope trace, I am not really seeing what is happening (good thing I never paid for the decoding, it is not so very useful)

So:
- Any chance Fremen you can, without too much effort, make a custom 5Mhz clocked version?
- Anyone has advice on what LA? Spec-wise I like the ScanaQuad SQ100, but its a bit pricey and I'd maybe like 8 instead of 4 channels.
  (maybe I should go and search a different thread for that, but here I am anyway)


PS: CHecked and re-checked, so pretty sure about the wiring. But if anyone wants to look, see pictures.
ALso, I keept the veroboard a little larger, so I can add buttons in the future...

There are lot of chances that it does not come from your bluepill board.

First test:  Just plug the 4 wires ribbon and you  should be able to talk to the board via PC Software.
If it works, second test: plug the two ribbons, except the /FPA Ready pin (you may have to de-solder one wire now with your vero board...). If it works then it comes from a different timing of this signal for whatever reason.. I could just release a FW version that continues init after a time-out. If it does not work than it might come from SPI init. I should have a look at how to recover from this one.
Depending on the results of those 2 tests, I could provide you with a special version with a lower speed SPI.

I will anyway add some diagnostic information in the next FW/PC Software release that could also help (provided that the serial link works with the 4 wires ribbon of course)

Regarding the LA, I would recommend the DsLogic Pro. I am very happy with it. Even the Basic version (like the one you mentioned in Germany) would allow live streaming @ 50Mhz and you could also upgrade it later on by adding a memory chip (when you feel confortable with SMD soldering...).

[cybermaus]

Back to the THS3095 topic.

My cheap Chinese "he promised they were genuine" THS3095 arrived and it was but a moments work to solder them in.
WOW, what a difference that made! I guess they are genuine. Or as good as.

Picture one: Yellow is 20Vpp into 50 Ohm with THS3095
Blue is reference waveform10Vpp into High-Z. You do not even see the blue because they are so equal.

To compare, The one from a few weeks ago, 20Vpp into 50 Ohm with THS3002
Do I need to say more?

[Marco]

Sellers on eBay and AliBaba that claim to be in UK or Germany are lying.
I know people do actually exist in UK and Germany, but only a few million, and there is like a few billion people in China, so statistically above statement is true.

I bought my FY6600 on ebay and it arrived within a week from a Chinese seller warehousing in Frankfurt and without custom headaches. I'm sure he doesn't live in Frankfurt, but he warehouses there.

Also, I noticed the eBay sellers do not include the special coax-probes, but just the straight lead probe

Them's the breaks. Don't see what coax probes are going to do on high impedance inputs any way. Proper high speed logic analyzers use attenuation probes which are terminated on the side of the analyzer pod. That's still the domain of more expensive devices though. If you want to you can twist two leads together and it accomplishes as much as that coax does AFAICS. Which is to say, bugger all.

[Candid]

Picture one: Yellow is 20Vpp into 50 Ohm with THS3095
Blue is reference waveform10Vpp into High-Z. You do not even see the blue because they are so equal.

Can you tell something about the reference waveform? Second channel into High-Z or what?

[cybermaus]

Yes indeed, just the B channel set to half the amplitude, and no 50 Ohm T at the oscilloscope.
Because there is no load, there is no trouble driving it for the OpAmp

I do not own a Spectrum Analyser, nor an actual reference, so this is the best I can do. So you may still start talking about -dB's and such, but just eyeballing the 3002 into 50 Ohm is so bad I do not need a SA or reference anyway.

And in any case, whatever error there is that I cannot see by eye, at least now the 50 Ohm is as good/bad as the High-Z, they overlap exactly.

[SMB784]

1. Power supply
Classical transformer 14Vx2/20VA
Linear regulators 14V DC for some output peak headroom - possibly LM317/337

Make sure the 14V has good good buffer caps, or else make the rail 12.5V instead of 14V. Because 14Vdc from 14Vac may cause some ripple if your caps are too small.
Especially since it was reported the 3095 on 11.5V already barely had clipping, so better a stable 12.5V then a ripply 14V

DC-DC to 6V and then linear to 5V

No need. That 5V is not used except to have a linear to 3.3V. So your DC-DC may as well go to 5V directly

Diode clipping to the power supply rails - BAV99

Ok, I missed that one. Why/How clip to rail with BAV99? Is that still needed if you replace the entire PSU?

So I have bought a Mean Well RD-3513 switch mode power supply with +/- 13.5V (tunable from 11.5V to 14V) output to replace the current power supply in the unit.  I'm assuming that removes the need for any transformers or associated regulators in the linear power supply solution quoted above.  If I use a lossy ferrite (any advice on the ferrite resistance?), in yall's estimation will there be any need for additional filtering to remove the switchmode noise besides the ferrite?  Will any other components besides the ferrite & new power supply be needed to power the dual THS3095 output op amps, other than a grounding cable to reference the BNC ground to mains earth?

Finally, are the THS3095's just a drop in replacement for the single output op amp or am I going to need to solder in additional components around them?

The upgrades that I will be performing will include swapping out the power supply, adding in a ferrite to suppress switchmode noise, swapping out the output op amps, grounding the BNC outputs, and installing a new TXCO reference (I went with the D75J).

Kammi, If this all goes as well as I hope it will, I will upload detailed step by step instructions to the github under the mods section with your permission.

[Candid]

Will any other components besides the ferrite & new power supply be needed to power the dual THS3095 output op amps, other than a grounding cable to reference the BNC ground to mains earth?

There is no need for a new psu for the THS3095s. A linear psu would prevent the ghost voltage to earth ground and keeps the outputs floating. Your reason for the new smpsu is because of the higher voltage? I replaced the smpsu with a self made linear one with 5V and +/-12V and do not have any problems with 20Vpp output on 50Ohm termination.

Will the mean well fit inside the housing?

Finally, are the THS3095's just a drop in replacement for the single output op amp or am I going to need to solder in additional components around them?

No additional components needed, just a replacement. I attach a picture.

...and installing a new TXCO reference

Soldering the TXCO modul is something you need patience for. I placed kapton tape on the bottom side of the module so that all the not needed pads were covered.

[SMB784]

Thanks, I needed the higher voltage supply to help handle the waveform clipping problem in the output. I just realized that I will need to put a 5V regulator after the 12v output in order to supply the necessary 5v for the rest of the board.  Any suggestions for how I should manage that?

[Candid]

For this you should take a step down converter that is capable of >1A not a 5V regulator like the 7805. On the 5V rail there is up to 1A needed. Using a 7805 it would need to dissipate (14V - 5V) * 1A = 9V * 1A = 9W! You get step down converters very cheap as ready to use modules. With a small heatsink the XL4015E1 (see picture) would be a good choice. It is capable of up to 5A. This module is only about 5cm x 2,5 cm x 2,5cm.

[DC1MC]

Thanks, I needed the higher voltage supply to help handle the waveform clipping problem in the output. I just realized that I will need to put a 5V regulator after the 12v output in order to supply the necessary 5v for the rest of the board.  Any suggestions for how I should manage that?

Personally, one these days I'll do the modifications with the nice 5EUR transformer and 3 linear regulators, plus 2 rectifier bridges and a 6 capacitors, but if you went on the SMPS way, then grab any Aliexpress/fleabay little 5V board and addit there, I think the smallest one was 2x3cm.

https://www.ebay.de/itm/12V-Step-Down-Modul-5V-3A-Konverter-Adapter-Power-Regler-Spannungswandler-MA1391/162762631499
https://www.ebay.de/itm/Micro-UBEC-5s-Lipo-23V-to-5V-6V-12V-Spannungswandler-Mini-FPV-BEC-5g/152707206751

My favorite trafo:
https://www.ebay.de/itm/Netztransformator-230V-8V-1-7A-4-2V-0-15A-13-6Volt-2x0-6A-Mittenanzapfung-NEU/152934971288

 Cheers,
 DC1MC

[cybermaus]

First test:  Just plug the 4 wires ribbon and you  should be able to talk to the board via PC Software.

confirmed: both the vero-PB and the loom-BP connect to PC software if only the 4-pin is connected.
btw: unlike what was stated before, it is the green LED that turns off after init.

If it works, second test: plug the two ribbons, except the /FPA Ready pin (you may have to de-solder one wire now with your vero board...). If it works then it comes from a different timing of this signal for whatever reason.. I could just release a FW version that continues init after a time-out. If it does not work than it might come from SPI init. I should have a look at how to recover from this one.

Both the vero-BP and the loom-BP do not work if I disconnect the /FPGARDY signal.

For added info, I attach a scope view of the SPI bus signals. Not sure if it helps in any way.
The first 2 are that I captured between BP and FPGA. The last 2 between FP and FPGA. While there are some overshoots on vertical edges, there is no noise, and there is a good clear difference between 0 and 1.

When BP is connected, these signals are shown over and over again. Hence the they are captured in RUN and nicely bright. My guess is the BP is trying to get the FPGA to say something, but it stays silent.
When FP is connected, the SPI is pretty much silent. I have to press a button to capture some bytes. Hence captured in SINGLE.

[fremen67]

@cybermaus: any luck with the bluepill now?

[cybermaus]

You may have missed the message just above yours.
Forgot to press refresh I guess.

Also, I ordered the DSLogic. From original website. No mention of delivery times, so likely 2 weeks minimum. But I am gone all next week anyway.

[bugi]

For this you should take a step down converter that is capable of >1A not a 5V regulator like the 7805. On the 5V rail there is up to 1A needed. Using a 7805 it would need to dissipate (14V - 5V) * 1A = 9V * 1A = 9W! You get step down converters very cheap as ready to use modules. With a small heatsink the XL4015E1 (see picture) would be a good choice. It is capable of up to 5A. This module is only about 5cm x 2,5 cm x 2,5cm.

On the list of DC-DC step-downs, here is what I got just yesterday (so I haven't tested them yet, EDIT: see below), but I had/have a somewhat similar need, though less critical on the noise - just need to get an aux voltage from 55V DC down to run a small fan. And a LED or two, a must have even with no real need/purpose, of course. If these work as advertised, certainly nice and tiny (just about 25.5mm x 16.5mm, looked so "big" in the photos, but in real life, smaller than the end of my thumb... and it is not even the smallest model I've found ).
https://www.ebay.co.uk/itm/DC-DC-Buck-Step-Down-Converter-48v-36v-24v-18v-12v-9v-to-3-3v-5v-2A-Power-Supply/252932129152
(May need to search a bit to find the same in other ebays, but IIRC, I first found that (or look-a-like) in the US-site.)

EDIT: checked it before powering up and the circuit seems to have some "why, oh why" -stuff:
Assuming the IC works as a real MP4560DN as it claims to be.  The chip's EN pin has been connected to the midpoint between board's edge EN--100k--*--39k--0V, and seller's instructions claim that the EN (at board edge) can be tied to the Vin (upto 55V) to enable output. Well, at 55V in, the midpoint would be at about 15.7V and datasheet says absolute max for EN pin is 5V. Uhoh.
Leaving the board's EN connection floating makes the midpoint pulled down, thus disabling the IC. (1µA pull down from EN-pin's internally driven 3.0V is enough, and the 39k certainly pulls plenty more). The worst thing is, it would have been easier to just bring the EN pin to board edge as is, and let the user pull it down for disable, instead now needing to drive it high (to a voltage between about 6V and 15V) to enable.
I guess I'll just remove the 39k resistor, which will make the EN pin "floating" and thus the IC by default enabled. Desoldering tiny SMD resistors, bleeeh...

[Marco]

Also, I noticed the eBay sellers do not include the special coax-probes, but just the straight lead probes.

Are there any sellers which advertise the Plus with coaxial leads? I've seen reviews of it, but all the pictures of the Plus on say aliexpress have 4+1 flying leads for each connector. The images for the plus on the Sigrok site do too. (And for some reason the DreamSourceLab website doesn't mention the plus at all.)

[fremen67]

You may have missed the message just above yours.
Forgot to press refresh I guess.

Oups! I was still on page 47 ...

Also, I ordered the DSLogic. From original website. No mention of delivery times, so likely 2 weeks minimum. But I am gone all next week anyway.

Nice choice.. plus,  should it happen, I would be able to read and analyze your BP captures 

First test:  Just plug the 4 wires ribbon and you  should be able to talk to the board via PC Software.

confirmed: both the vero-PB and the loom-BP connect to PC software if only the 4-pin is connected.
btw: unlike what was stated before, it is the green LED that turns off after init.

If it works, second test: plug the two ribbons, except the /FPA Ready pin (you may have to de-solder one wire now with your vero board...). If it works then it comes from a different timing of this signal for whatever reason.. I could just release a FW version that continues init after a time-out. If it does not work than it might come from SPI init. I should have a look at how to recover from this one.

Both the vero-BP and the loom-BP do not work if I disconnect the /FPGARDY signal.

For added info, I attach a scope view of the SPI bus signals. Not sure if it helps in any way.
The first 2 are that I captured between BP and FPGA. The last 2 between FP and FPGA. While there are some overshoots on vertical edges, there is no noise, and there is a good clear difference between 0 and 1.

When BP is connected, these signals are shown over and over again. Hence the they are captured in RUN and nicely bright. My guess is the BP is trying to get the FPGA to say something, but it stays silent.
When FP is connected, the SPI is pretty much silent. I have to press a button to capture some bytes. Hence captured in SINGLE.

OK. I think I got it. In fact it should be stucked during flash read tests. As I already added a 2s time-out during FPGA init, it should have failed to initialize the FPGA, went on after 2 sec and jumped to flash eeprom selftest. The FPGA not beeing initialized, it can't answer to flash read requests. That would explain what you see on the scope.
So I have to:
- add time-out to FPGA requests anyway
- skip flash read tests when failling to initialize FPGA
- Add more diagnostic data to startup phase.
I did not spend a lot of of energy in startup sequence diag but I will do it now.

Still it does not explain why the FPGA could not be initialized...
Now the LA would have been indeed handy.

Did you already tried to flash your eeprom with v3.2 ?
Could be interresting to see if it's related to v3.1

Edit: There should be something else . If the FPGA is not initialized at all, it can't block the startup sequence (the case when only connecting the 4 wires ribbon...)