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

[DaveR]

 No sense breaking something that works until you know the replacement works.

Exactly.  As you seem to be the only one with the luxury of having a spare front panel to play with, Reg, we'll be counting on you for firmware testing.  There are a few more of us to debug the control software and blue pill board, without risk of uncovering a new "3.0" type glitch and ending up with a brick which needs delicate chip replacement surgery .

Regards,
Dave

[DaveR]

Hi fremen67:

I tried to start the software testing today, but there appears to be a problem with the 0.8 blue pill hex file - it loads ok, but when the blue pill board is plugged into a USB port the PC throws up a "Device not recognised - a device ..... has malfunctioned)" message.  I can reload 0.6 and 0.7 hex files on the same board and they'll work fine, but 0.8 causes the error message again.  I tried four different boards and it's the same with each.  It wouldn't have anything to do with 0.8 being 3x the size of 0.7 would it (it does seem to be a very large jump in size compared with previous versions)?

Regards,
Dave

[fremen67]

Hi fremen67:

I tried to start the software testing today, but there appears to be a problem with the 0.8 blue pill hex file - it loads ok, but when the blue pill board is plugged into a USB port the PC throws up a "Device not recognised - a device ..... has malfunctioned)" message.  I can reload 0.6 and 0.7 hex files on the same board and they'll work fine, but 0.8 causes the error message again.  I tried four different boards and it's the same with each.  It wouldn't have anything to do with 0.8 being 3x the size of 0.7 would it (it does seem to be a very large jump in size compared with previous versions)?

Regards,
Dave

This is completely normal. The code of the bluepill is exactly the same as the front panel version, the difference is that some I/Os of the front panel can't be used on the bluepill and are remapped for it. But if you connect them, it will handle the LCD, the keyboard and the encoder.
Almost all the I/Os of the bluepill are used now, including PA11 which is USB- signal, now used as an output to switch on the LCD.
For the test, you don't need to connect anything on the USB port of the bluepill.
Just connect the bluepill the way it was with v0.7 and it should work like this. Did you try?

Edit: just a picture to show what you could do with the current bluepill  firmware...

[DaveR]

Just connect the bluepill the way it was with v0.7 and it should work like this. Did you try?

Yes, I did - I even reprogrammed the blue pill without disconnecting it from the FY6600 to make sure everything was the same for each BP level, but the v0.8 control software would do little or nothing.  However, I've now tried it again after rebooting the PC, and without the myriad of SDR radio programs and other stuff running in the background, and everything is working properly this time - so it's all systems go! (And now I know why it was throwing up the error, as the BP is effectively no longer a USB device with the latest firmware loaded.)

I've just given the Counter function a workout, and it's fine with AC coupling - within the limits of my FY6600, anyway.  The manual claims that the counter can measure up to 100MHz, but I remember from when I last tried it several months ago that mine wouldn't go anywhere near that.  Having tried it just now, it will only go up to 38.1MHz with a sine wave input, and 42.5MHz with a square wave, so I don't know whether this is just a problem with my unit or with the counter in general.  (That's actually a lot better than when running with FP control, when it falls over above 29.7MHz with any input.)  Has DC coupling via the Trig IN port been enabled yet?  I can't get any response after I select it, although it's hard to say whether it works under FP control anyway - it never shows a frequency, and the values it does show for Cycle, Width and Duty only rarely coincide with actual values.

A lot more to be done tomorrow, starting with the Calibration routines ....

Regards,
Dave

[soundtec]

Tremendous work Fremen,

In the case where the front panel is reprogrammed with your software will the internal usb provide the software functionality also , or is the bluepill controller still required ?

The new simplified gui looks clearer and  easier to read ,as well as giving  controls well beyond the standard FP setup ,
From purely a visual perspective I felt a little more curve at the edges of the boxes and soft key tabs on the screen might be more pleasing on the eye , thats all only icing on the cake in the end,

I have two working units , so I'm strongly weighing the options and considering rebranding my modded  fy6600 'Frementec'.
Does anyone know where the idiots guide to the programmer hook up is located at ?

[fremen67]

Has DC coupling via the Trig IN port been enabled yet?  I can't get any response after I select it, although it's hard to say whether it works under FP control anyway - it never shows a frequency, and the values it does show for Cycle, Width and Duty only rarely coincide with actual values.

A lot more to be done tomorrow, starting with the Calibration routines ....

Regards,
Dave

Yes the Trig IN counter works as well. I reached up to 104Mhz in the front input and 120 Mhz in the Trig IN input with a sine signal.

For measurement, the main work is done by the FPGA. The FP tells the FPGA what input to use and then checks on a time base when a value is available for reading.
The FP then gets Frequency or Counter depending on mode, gets also +/-WIDE from the FPGA and then do the maths for Period and Duty values.

[fremen67]

In the case where the front panel is reprogrammed with your software will the internal usb provide the software functionality also , or is the bluepill controller still required ?

You don't need the Bluepill when you reflash the front panel. And yes the USB connection on the rear of the FY6600 will work when connected to a PC with the PC Software.

The new simplified gui looks clearer and  easier to read ,as well as giving  controls well beyond the standard FP setup ,
From purely a visual perspective I felt a little more curve at the edges of the boxes and soft key tabs on the screen might be more pleasing on the eye , thats all only icing on the cake in the end,

I also like the curve at the edge of the boxes but I wanted the UI to be as fast as possible and drawing horizontal and vertical lines goes faster then drawing arcs… I will keep that for the end 

I have two working units , so I'm strongly weighing the options and considering rebranding my modded  fy6600 'Frementec'.
Does anyone know where the idiots guide to the programmer hook up is located at ?

That would be very nice! What versions do you have?

As for flashing the front panel, it is a matter of soldering a 4 pins header to JK1 (just under the 8 wires ribbon) and wiring  GND, JTCK and JMS to a ST-Link V2 programmer.
JTCK to SWCLK, JMS to SWDIO and GND … to GND. Keep the 2 ribbons connected as the power supply will come from them.
In short you switch on the FY6600 , launch the STM32-utility, click Target/Program&Verify,  select the .hex file and start!
You will find the software for the ST-Link here: https://my.st.com/content/my_st_com/en/products/development-tools/software-development-tools/stm32-software-development-tools/stm32-utilities/stsw-link009.html

[DaveR]

Yes the Trig IN counter works as well. I reached up to 104Mhz in the front input and 120 Mhz in the Trig IN input with a sine signal.

I've worked out what the problem is now: the counter is not very sensitive and needs a fairly high input voltage, so if you feed the counter from one of the output channels the voltage will drop off as the frequency is turned up, and soon falls below the counter's threshold level.  The display and the Manual both warn about not exceeding 5V at the counter inputs, but what they don't state is that the level shouldn't fall too far below that, either, so the FY6600 effectively can't measure itself above about 30MHz.

It's been a bad day for testing, I'm afraid, as endless "events" stopped me from even getting started.  With luck I'll be able to stay clear of distractions tomorrow and get some good time in.

Regards,
Dave

[soundtec]

I have a version 3.2 15mhz ,

I have the St-32 programmer and software to hand  , I just need to solder in header pins and I can try running it up on the new UI  later today .

[soundtec]

That flashed  fine for me , its up and running on 0.8 ,
PC Software clicked in nicely too ,
Only unexpected thing I found was pressing the encoder button defaulted  frequency to 10000hz and no longer changed  hz/khz/mhz .
I only gave it a quick visual check over so far ,I'll have another closer look in the afternoon .


 

[soundtec]

I discovered the way the left right arrow buttons work while in wave select mode , to change bank from preset to arb thats a handy feature.

Everything just feels more snappy now with the new code , its not like the original s/w where the machine seems to have a little yawn before doing what its told , same can be said for the new pc front end software , when you change something on the pc screen theres no decernable delay in whats taking place in the unit itself.

One thing I've never liked much about the F'tec was the need to  press the channel button twice  in order to mute or switch on the output channels , if channel select(to top of screen)  and on or off could happen in a single key stroke it would be much handier for cueing and mixing the outputs for test signals or even if the unit was used in a musical context . 

So apart from the frequency select mode issue on the encoder push button  and the extra zero's that appear in front of the selected field which I thought were unessesary ,I cant fault one single other thing in the new UI , it has a definate edge in terms of intuitiveness of usage compared to Ftec 3.2 , I also like the way the parameters visible and available to modify changed depending on the waveform in use , it helped me focus better on the job at hand .

The new measure ,mod and sweep pages are a revelation compared to 3.2 , and full independent mod/sw  features on both channels  not to mention  my formerly 15mhz machine now goes up to 62.5mhz   
you have truly spoiled us sir !

The firmware  upgrade was easy , Id had the St32 programmer software on my pc before so once I got the header implanted,3 wires  and the driver clicked  in  ,once Id ok'd overwrite warnings , took about 30 seconds ,the machine rebooted itself , and it runs off the new UI without a glitch since. 

[MikeLud]

Fremen67,

Do you have an estimate when you will have a version that will work with a FY6800. I have a 60mhz FY6800 board version 1.7 FW V1.7.1. If you need help testing let me know. Also what LCD are you using with the bluepill.

Thanks
Mike

[jleg]

As for flashing the front panel, it is a matter of soldering a 4 pins header to JK1 (just under the 8 wires ribbon) and wiring  GND, JTCK and JMS to a ST-Link V2 programmer.
JTCK to SWCLK, JMS to SWDIO and GND … to GND. Keep the 2 ribbons connected as the power supply will come from them.
In short you switch on the FY6600 , launch the STM32-utility, click Target/Program&Verify,  select the .hex file and start!
You will find the software for the ST-Link here: https://my.st.com/content/my_st_com/en/products/development-tools/software-development-tools/stm32-software-development-tools/stm32-utilities/stsw-link009.html

I just flashed my bricked V3.0 - worked like a charm! Thank you very much for these efforts!
Still didn't get anything from Feeltech - i owe you a beer at least 

Here's what i had before your cure, and what i have now:

[fremen67]

I discovered the way the left right arrow buttons work while in wave select mode , to change bank from preset to arb thats a handy feature.

I completely forgot to describe the new UI funtionnalities but it seems that you discovered them all by yourself...
For the others who do no have it in front of them:

Waves:
- First click on Wave button change to edit mode but does not change the current selection (big pain with the original firmare ...)
- Encoder +: change current wave to the next one
- Encoder -: idem for previous one
- Right Arrow: go to Arb1 wave
- Left Arrow: go to Sine vave
- Encoder click: end edit mode

Numerical parameters (frequency, ampl, ...):
- First click on corresponding function key change to edit mode
- Right, left arrows: select next or previous digit
- Encoder +/-: inc/dec selected digit
- Encoder click: select default parameter value (0V for Offset, 50% for Duty, 0° for phase, etc ...)
- Second click on corresponding function key end edit mode

Text parameters (sweep source, source direction, etc ....):
- First click on corresponding function key change to edit mode
- Right, left arrows, Encoder +/- : select next or previous choice
- Second click on corresponding function key: show possible choices for parameter on function keys for direct selection
- Encoder click: end edit mode

One thing I've never liked much about the F'tec was the need to  press the channel button twice  in order to mute or switch on the output channels , if channel select(to top of screen)  and on or off could happen in a single key stroke it would be much handier for cueing and mixing the outputs for test signals or even if the unit was used in a musical context . 

OK I will do it!

So apart from the frequency select mode issue on the encoder push button  and the extra zero's that appear in front of the selected field which I thought were unessesary ,I cant fault one single other thing in the new UI , it has a definate edge in terms of intuitiveness of usage compared to Ftec 3.2 .

I will change the behaviour of the encoder click when on frequency (revert to defaut value at the moment). It's in my todo list, before the rounded corners but after the system menu 
As for the leading zeros appearing on edit mode, it has been done on purpose to ease digit selection but other solutions could be used later on (like adding zeros only to the right of the cursor).

The new measure ,mod and sweep pages are a revelation compared to 3.2 , and full independent mod/sw  features on both channels  not to mention  my formerly 15mhz machine now goes up to 62.5mhz   
you have truly spoiled us sir !

Thanks a lot for the compliments  (Sweep on both channels but modulation on channel 1 only as it is done by the FPGA...)

Some days ago I was on the idea to release only the bluepill version to test the functionnalities and now the front panel version is already installed on some units... I even did not finished to test all the programmed GUI functions.
You are the Indiana Jones of the FY6600 ! No fear!

I found and corrected yesterday 2 bugs in the version you have. The counter reset function did not work anymore and the encoder driver had some issues when activating a sweep function.
I will give you another version in some days.

[fremen67]

Fremen67,

Do you have an estimate when you will have a version that will work with a FY6800. I have a 60mhz FY6800 board version 1.7 FW V1.7.1. If you need help testing let me know. Also what LCD are you using with the bluepill.

Thanks
Mike

Hi Mike,
I have to focus on finishing the remaining menus (at least system menus) before switching to FY6800. The first step will be to analyze the protocol between the front panel and the FPGA to see the differences with the FY6600. I suppose that it will not impact a lot the new firmware (this plus some different keys on the 6800 front panel) but I can't be sure now. I will say one or two months... I will anyway do the modifications for the 6800 and the 6900 in parallel to same time.

The new firmware will be a big improvement for the sweep function of the FY6800 as they ruined it. Try this:
connect an oscilloscope and sweep a 0V DC wave from -1V to +1V at 0.01s or even at 0.1s sweep time. The ramp obtained is a disaster on my 6800 which is V1.6. May be different on the V1.7 but I have doubts.

In the meantime if you have a way to backup the flash eeprom on the main board that would be interesting for me.
The LCD that would work for the bluepill is a ILI9341 with SPI.

[fremen67]

I just flashed my bricked V3.0 - worked like a charm! Thank you very much for these efforts!
Still didn't get anything from Feeltech - i owe you a beer at least 

Thank a lot for the feedback!
It is still in development but what about beeing involved in testing the firmware? 

[fisafisa]

Hi
Very good job!
Been reading for a long time until you disappeared.
Now I am very happy to see the project back online.

Any chance of releasing the source code?
I was mulling about porting it over to either a F7 discovery board I have with touch screen
or directly to a Raspbeery pi with a large screen.

In any case
Thanks

[soundtec]

I spotted some other slightly odd behaviour relating to the gate setting  and displayed frequency in measure mode , this also seemed to be having some effect on the pc softwares  stability. I also couldnt seem to get anything from the trig input on dc mode .

Please dont feel any need to rush an update on my behalf , all in its own good time will be fine ,
I have a second machine so Im quite happy to play guinea pig with the other .

[Miti]

It doesn't work with a V3.3 eeprom but as I don't have a V3.3 front panel, I can't see what changed in the FP-FPGA protocol for v3.3.

Hi Fremen67 and welcome back! I surely hope that you were just busy and nothing tragic happened.

How may I help to figure out V3.3 and what do I need? My FY6600 is V3.3. I don't have a logic analyzer but I do have a Rigol DS1054Z. Probably not enough, eh?

[DaveR]

I spotted some other slightly odd behaviour relating to the gate setting  and displayed frequency in measure mode , this also seemed to be having some effect on the pc softwares  stability. I also couldnt seem to get anything from the trig input on dc mode .

Hi soundtec - it's getting to be just like the good old days again, isn't it?

Try this and see if you can get your Trig IN working:  use your new FremTech 6600 box to send a 20MHz sine wave to the Trig IN of your 6800, and set the amplitude to just under 5v.  It should be measured ok on the 6800 counter with DC coupling.  Now increase the frequency to 40MHz, and the 6800 counter will probably not measure it.  Increase the amplitude until the relays activate and the 6800 will measure the signal again.  Go to 50MHz and the amplitude will need to be set to about 9.1v, and at 54MHz it will need to be about 18.7V.  Mine tops out at just below 56MHz, when the 20v amplitude limit is reached (although the RMS output level is still less than 5v, so still within the stated safe input limit on the 6800).  If you've changed the op amps on your 6800 you should be able to control it with fremen's software and test the Trig IN port on the FremTech (I can't because I still haven't changed the op amps in the 6800, so can't get anywhere near enough amplitude on the signal to activate the counter at anything over about 10MHz).

If you use the front input on the 6800 you should be able to send a two-cycle sine wave from the FremTech and check the counter up to >90MHz.  I've just done this on mine, and went up to 93MHz with 17.15v amplitude set (only 6.2v pp out, 2.1v RMS), so I could probably have got another 3 or 4 MHZ out of it before I hit the 20v limit.  Fremen must have a signal generator which can produce high RMS outputs to push the counter as far as he did.

What irregularities did you spot with the gate and software stability?  I'll see if I can replicate the problem here if you give me some settings to enter.

Regards,
Dave

[soundtec]

Hi Dave ,

With a 1khz 5v square wave output to
dc/trig or ac  input to measure on the Frementec
Frequency display changes depending on which gate setting I use ,on 1S reads good ,10S reads 10khz  ,on 100S mode reads as 100khz. Also at power up gate reads 'Mode' not 1S 10S or 100S.  I just did the same test on Fy6800 and it properly displays 1khz on all gate settings .

My earlier issue with the Trig input was down to the patch cable .

While making adjustments to the measure panel on the pc I did get a couple of crashes where the program just closes after an error message .

[DaveR]

Were you seeing the wrong values on the front panel or the PC?  I haven't flashed my FP yet, but the PC software shows 1.000kHz, 1.0000kHz and 1.00000kHz, which is as you'd expect.

I've only managed to find one glitch so far, and that's on the Calibration Panel, when it does the same error behaviour.  It's looking pretty resilient and stable otherwise.

[soundtec]

Theres a few small changes required for the FP software to be working right , I believe the bluepill software is a little ahead in that respect . Fp adds the extra 0 like you see Dave ,but looks like it forgets the decimal point ,so the number looks wrong .

 I have an idea which might be a nice feature to add,

The abillity to arrange and automate a sequence of tones or arb waveforms  , that would be incredibly usefull for test purposes .
Rs232 is available ,would another usb to serial link connected to backpanel tx rx and gnd be the best way to go that way its entirely seperate from the internal usb used for the pc software . 

https://www.partsnotincluded.com/midi/how-to-send-receive-messages-over-serial/
that might do it

[jleg]

I just flashed my bricked V3.0 - worked like a charm! Thank you very much for these efforts!
Still didn't get anything from Feeltech - i owe you a beer at least 

Thank a lot for the feedback!
It is still in development but what about beeing involved in testing the firmware? 

sure, hopefully this weekend i'll get to this, and try to dig a bit into testing...

[soundtec]

I loaded up the fy6600 serial protocall in open office calc ,seems to display nice ,but I havent a clue how it works ,

I saw the link to the guy who made WSPR the transmitter from the fy6600 ,
It seems like a way to automate the output of the generator would be really usefull for radio
http://www.arrl.org/files/file/QEX_Next_Issue/May-June2019/Steber.pdf
In this case a way to transcribe a text message into a waveform which is then uploaded to an empty slot in the Feeltech and schedualed for transmission at the next time interval,  all automatically ,

I did have some limited success with the with the serial to midi  I mentioned in my last post  ,
I was able  recieve a signal after power  on from the Feeltech via midi in my daw , but I dont know how to send a meaningfull control signal the other direction .  The idea of being able to use the feeltech in conjunction with midi and a sequencer seems to appeal to others as well .
For audio testing the excellent 'Room Eq Wizard' by John Mulcahy has a midi and serial connections tab , if it had the abillity to control a  generators output via serial  that would be nice .

I get the impression everything in the above wish lists is easily possible , all the hardware is there , the control protocall is there in black and white , its more like just hooking up the numbers in software .

https://www.eevblog.com/forum/testgear/feeltech-fy3224s-24mhz-2-channel-dds-aw-function-signal-generator/550/

He seems to have made his pwn dedicated remote