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

[cybermaus]

There was this (I think) russian guy who said he was familiar with this FPGA and that it did have some possibly better clock options. But we have no source for the FPGA, so it would mean re-implement the entire FPGA firmware from scratch. Kind of like what Fremen67 is doing for the CPU. But that will be a massive undertaking so not likely someone with the right skillset picks that up.

Practically, I think you should accept the 4ns thing as a given fact.

[Gandalf_Sr]

... I have also always been frugal (aka a cheap bastard) and am still amazed at myself how I buy a cheap Chinese AWG when a) I can easily afford the proper stuff and b) I seem to spend relatively high amounts on patching up said AWG with PSU, OpAmps, etc. Truth is, I think I enjoy the challenge and frustration of making things work for pennies on the dollar.

I'm in the same category, I have a large box full of such half-assed projects but, this time, I was able to talk myself out of buying yet another prototype.  When I first saw this thread I thought "wow, a dual AWG for $100 but the additional cost in 'upgrades', shipping, and time probably triples that price.

If you want to generate dual arbitrary waveforms and have a good range of built-in useful waveforms (e.g. ECG), buy a debugged, working product like a Rigol DG1022Z (which I did for around $325.)

If you want to spend about the same amount of money (in the end), waste hours tinkering, changing ICs, waiting 3 weeks for cheap programmers to arrive from China, and end up with something that sort of does the job, buy the FY6600/FY6800.

Your choice.

[DC1MC]

I'm in the same category, I have a large box full of such half-assed projects but, this time, I was able to talk myself out of buying yet another prototype.  When I first saw this thread I thought "wow, a dual AWG for $100 but the additional cost in 'upgrades', shipping, and time probably triples that price.

If you want to generate dual arbitrary waveforms and have a good range of built-in useful waveforms (e.g. ECG), buy a debugged, working product like a Rigol DG1022Z (which I did for around $325.)

If you want to spend about the same amount of money (in the end), waste hours tinkering, changing ICs, waiting 3 weeks for cheap programmers to arrive from China, and end up with something that sort of does the job, buy the FY6600/FY6800.

Your choice.

Uhmm, a 60MHz Rigol is SEVEN times more expensive compared with the equivalent Feeltech, it has a sampling rate of 200Ms/s instead of 250Ms/s and in the end is still a Chinese product.
Also it doesn't have our Freemen excellent software that is getting better by the day  . and you have to put a lot of money in the linear PSU and op amps to gome even closer to 25% of the RIgol.

 Cheers,
 DC1MC



 

[DaveR]

If you are going to quote price comparisons, Gandalf, at least give like for like, eg at current UK prices:-

Feeltech FY6600 (60MHz) - £85

RIGOL DG1062Z (60MHz) - £654

My upgrades have cost me about £50 on top of the £85, so £135 in total: still only just over 20% of the cost of the RIGOL.  Yes, the RIGOL has a some extra features, but is it really worth five times as much??

If my FY6600 dies on me I can buy another and transfer the mods in an hour for nothing.  I'm happy with that, and so, I would imagine, are most of the users on here.

[cybermaus]

If you want to generate dual arbitrary waveforms and have a good range of built-in useful waveforms (e.g. ECG), buy a debugged, working product like a Rigol DG1022Z (which I did for around $325.)

What I would find really interesting (not even out of spite or to counter arguments, but really I am interested to know) is if your Rigol also has this base-clock-resolution jitter problem. In other words, does your 200MS/s Rigol show 5ns jitter on vertical edges.

It may be a bit harder to tell, as you only have a 25MHz device, and you'd have to check on a non-dividable frequency, so not 20MHz or 25MHz but maybe 22MHz.
Do you see anything like the yellow wave in this post, but with 5ns distance instead of 4ns, on square waves or arbitrary waves with hard vertical edges?

[nctnico]

... I have also always been frugal (aka a cheap bastard) and am still amazed at myself how I buy a cheap Chinese AWG when a) I can easily afford the proper stuff and b) I seem to spend relatively high amounts on patching up said AWG with PSU, OpAmps, etc. Truth is, I think I enjoy the challenge and frustration of making things work for pennies on the dollar.

I'm in the same category, I have a large box full of such half-assed projects but, this time, I was able to talk myself out of buying yet another prototype.  When I first saw this thread I thought "wow, a dual AWG for $100 but the additional cost in 'upgrades', shipping, and time probably triples that price.

If you want to generate dual arbitrary waveforms and have a good range of built-in useful waveforms (e.g. ECG), buy a debugged, working product like a Rigol DG1022Z (which I did for around $325.)

If you want to spend about the same amount of money (in the end), waste hours tinkering, changing ICs, waiting 3 weeks for cheap programmers to arrive from China, and end up with something that sort of does the job, buy the FY6600/FY6800.

Your choice.

This isn't an entirely accurate statement. The Feeltech generators will do fine as a basic function generator. For many tasks you don't really need a generator with many bells and whistles so the Rigol and Siglent generators (which aren't free of flaws) are just not worth the money if all you need are basic waveforms.
However some go way overboard by fitting a completely new PSU and so on in their Feeltech generator turning it into a neat tinkering project. In the Feeltech FY3200S I have I fitted a grounded mains inlet and a wire from the earth connection of the PSU to earth to get rid of the leakage problem. That was a minor amount of work. The FY6800 seems to be a newer design with a grounded mains inlet so it is different compared to the previous generators. I guess we'll be seeing an FY6800 teardown pop-up in the near future.

[kahe40]

... some go way overboard by fitting a completely new PSU and so on ...

really funny, each time they connect the BNC-cable to their scope, or USB-cable to their PC,
it is grounded - gratis - no need for nothing.

[Mr. Scram]

really funny, each time they connect the BNC-cable to their scope, or USB-cable to their PC,
it is grounded - gratis - no need for nothing.

Were those roughly 60 volts on the neutral part of the BNC on the generator imaginary too?

[nctnico]

... some go way overboard by fitting a completely new PSU and so on ...

really funny, each time they connect the BNC-cable to their scope, or USB-cable to their PC,
it is grounded - gratis - no need for nothing.

The problem is that until it is connected to ground you have 60V to 120V AC floating around and that will kill your circuits for sure.

[cybermaus]

The problem is that until it is connected to ground you have 60V to 120V AC floating around and that will kill your circuits for sure.

That is a bit of an exaggeration. It is only capable of a few uA, and may kill the most sensitive of test circuits, but it is not a "sure" thing.

Any 2-prong power supply has this problem, it is not necessarily due to bad construction, but caused by the regulatory required Y Cap. They all will float in the middle at zero load, but even adding a DMM probe will pull it down from 115V (middle of 230V) to 60V or 40V or less, depending on how big a Y-cap and how much impedance your DMM has.

That is not to say I like this, hence I also added a 3 prong, rerouted the Y cap to that, and put on a 1M preload. But saying "kill for sure" is a bit too harsh a statement. That's like saying your USB chargers "for sure" kill all your devices (yes, most USB chargers also have this)

[soundtec]

I think if you've gone with the op amp upgrade its well worth doing something about more current on the +/- rails ,a pair of 3095's pull twice the current of the dual channel 30021 that its replacing . As far as using all three pads with two single channel and one dual op amp it would be do-able but external output mixing resistors for each op amp would have to be arranged and the final output padding resistor would need to be adjusted. With the new TI op amp  being able to draw 350ma paralleling op amps hardly seems nessesary unless you want super low load drive.

Still waiting on the 6800 and what ever surprises it brings ,
I might try a 1:1 isolation transformer on this chassis ,and maybe two seperate switched rails for 5V and +/- .

Dave's right , for the 55 euro's it cost (15mhz model)its well up with competition ,and the extra spend on psu is well worth it,the time ,well it depends on how much you have to hand

I'm having a look at different proprietry software relating to AWG's ,its all a bit samey,B+K Wave express ,Ultrawave (Rigol),Arbexpress(Tek) and Easywave (Siglent).

I found this ,Im not sure if its related to Feeltech or not ,I also cant figure out what its supposed to be doing,
https://www.spooky2.com/newmodel/
Does look a bit like the fy2300 though, has someone reinvented it as a 'Tens' machine  or something ? theres all kinds of extravagant claims on the website about it being used for cancer treatment ,
Any voulenteers for some pwm alphawave experiments? a few crock clips and leads should do it 

[Gregg]

First I would like to thank all of those who have actively participated in his thread and contributed toward making the FY6600 great (sometime, maybe). So far, it has been a great learning experience enabling my cheapskate self to possess a bargain signal generator. 
The last time eBay offered a 15% off coupon, I sprung for a US warehoused 60 MHz that had the latest firmware.  I had been eyeing them for some time, thanks to this thread.
I’ll have to post pictures when I get it back together after installing the THS3095 op amps and a better heat sink.  I managed to get everything in the case in a layout similar to the picture Dave posted including a 40mm 12v fan driven off the second 9v transformer winding.
 
Parts & Cost list:
$87.46 FY6600-60 eBay (with 15% off coupon) US stock V3.2.1
$21.08 30va R-core 18v 18v 9v 9v Transformer eBay (with 15% off coupon)
$6.99 LM317 LM337 ac to dc regulator kit eBay
$1.58 LM317 ac to dc regulator kit eBay
$7.66 two THS3095 op-amps eBay
$124.77 Total
40mm fan and rectifier, capacitor, resistor from my stock
C14 power inlet, fuse holder and fuse from my stock
Fasteners, stand offs, wire, etc. all from my stock

[soundtec]

Theres no doubt about it for the budget concious who dont mind doing a few uprgades the FY is a great deal .
It would be nice to see a like for like waveform comparison with the equivalent 60mhz Rigol unit and see if theres a difference .

[Gandalf_Sr]

First I would like to thank all of those who have actively participated in his thread and contributed toward making the FY6600 great (sometime, maybe). So far, it has been a great learning experience enabling my cheapskate self to possess a bargain signal generator. 
The last time eBay offered a 15% off coupon, I sprung for a US warehoused 60 MHz that had the latest firmware.  I had been eyeing them for some time, thanks to this thread.
I’ll have to post pictures when I get it back together after installing the THS3095 op amps and a better heat sink.  I managed to get everything in the case in a layout similar to the picture Dave posted including a 40mm 12v fan driven off the second 9v transformer winding.
 
Parts & Cost list:
$87.46 FY6600-60 eBay (with 15% off coupon) US stock V3.2.1
$21.08 30va R-core 18v 18v 9v 9v Transformer eBay (with 15% off coupon)
$6.99 LM317 LM337 ac to dc regulator kit eBay
$1.58 LM317 ac to dc regulator kit eBay
$7.66 two THS3095 op-amps eBay
$124.77 Total
40mm fan and rectifier, capacitor, resistor from my stock
C14 power inlet, fuse holder and fuse from my stock
Fasteners, stand offs, wire, etc. all from my stock

Time: 10 hours at $130 per hour = $1,300.00
Shipping for all the above (est)    =     $50.00
Cost of parts                              =  $127.44
Total                                          = $1,477.44

[bugi]

First I would like to thank all of those who have actively participated in his thread and contributed toward making the FY6600 great (sometime, maybe). So far, it has been a great learning experience enabling my cheapskate self to possess a bargain signal generator. 
The last time eBay offered a 15% off coupon, I sprung for a US warehoused 60 MHz that had the latest firmware.  I had been eyeing them for some time, thanks to this thread.
I’ll have to post pictures when I get it back together after installing the THS3095 op amps and a better heat sink.  I managed to get everything in the case in a layout similar to the picture Dave posted including a 40mm 12v fan driven off the second 9v transformer winding.
 
Parts & Cost list:
$87.46 FY6600-60 eBay (with 15% off coupon) US stock V3.2.1
$21.08 30va R-core 18v 18v 9v 9v Transformer eBay (with 15% off coupon)
$6.99 LM317 LM337 ac to dc regulator kit eBay
$1.58 LM317 ac to dc regulator kit eBay
$7.66 two THS3095 op-amps eBay
$124.77 Total
40mm fan and rectifier, capacitor, resistor from my stock
C14 power inlet, fuse holder and fuse from my stock
Fasteners, stand offs, wire, etc. all from my stock

Time: 10 hours at $130 per hour = $1,300.00
Shipping for all the above (est)    =     $50.00
Cost of parts                              =  $127.44
Total                                          = $1,477.44

That $130/h depends on where one lives, what one does for work (or if one even has a work), etc., and is the price one would perhaps pay to some company to do it, less than that if it was someone being paid personally. When one does it by oneself, the cost of time should be the amount one would get less income, considering taxes and all.. Typically closer to $15/h   If I'd do just the main recommended changes as is, without more thinking, the total time would be somewhere between 3-5 hours (+perhaps 1 hour for cleaning up).  And of course, can you really put a "cost" on free time, especially if what needs to be done is the same stuff you'd do anyway as a hobby.  But, lets give it a rounded $75 worth (of my time).

E.g. a typical student hobbyist with no work, that "work time cost" would be zero.

Shipping for parts, in my case (on the competing product, but basically corresponding changes), has so far been zero. I even had to order another shipment after realizing a mistake in my first design, and none of the orders were from ebay with its flood of free shipping parts (mostly digikey, couple local stores). Still zero... Of course, the shipment costs are baked into the part costs, but the part costs listed already are close enough what I have been paying, so I'd assume those prices also include shipping, or at least could include.  Granted, I did order other stuff at the same time, pushing the order totals above the free shipment limit, but I'd think most hobbyist can and would do so, too.  (Third order/shipment is possibly on the works, but I have to test some things with the already acquired parts before choosing what (or if anything) I will still need. I have so much other parts in my "to buy"-queue that all my digikey orders will be free for the foreseeable future.)


So, for me, the total would be around $200 even with the "work cost" (though I would really consider it a hobby project, so that $125 is better, with zero shipping cost). As it would be for large share of students and all unemployed people, even if they would not consider it a hobby.

That nearly $1.500 estimate is just nonsense. The only case where it would apply is if one is doing just one unit in a company, spending workers' time for it. No sane company would do that, so it would be nonsense either way . And even then, it is actually not the "cost", but "cost + less potential income" - a difference, which, it seems, 99% of company leaders do not understand.

[SMB784]

Time: 10 hours at $130 per hour = $1,300.00
Shipping for all the above (est)    =     $50.00
Cost of parts                              =  $127.44
Total                                          = $1,477.44

Your salary is $270,000 dollars a year? Where exactly do you come up with $130 per hour?

[Mr. Scram]

Time: 10 hours at $130 per hour = $1,300.00
Shipping for all the above (est)    =     $50.00
Cost of parts                              =  $127.44
Total                                          = $1,477.44

While I mostly agree with your reasoning, it would also mean I couldn't pick up a book or learn a new skill any more. I couldn't afford either.

[Gregg]

Parts & Cost list:
$87.46 FY6600-60 eBay (with 15% off coupon) US stock V3.2.1
$21.08 30va R-core 18v 18v 9v 9v Transformer eBay (with 15% off coupon)
$6.99 LM317 LM337 ac to dc regulator kit eBay
$1.58 LM317 ac to dc regulator kit eBay
$7.66 two THS3095 op-amps eBay
$124.77 Total
40mm fan and rectifier, capacitor, resistor from my stock
C14 power inlet, fuse holder and fuse from my stock
Fasteners, stand offs, wire, etc. all from my stock

FYI: Shipping was included in the above list.
I am retired and can get part time consulting jobs that pay more than $130/hr they are few and far between; the upside is I get to pick and choose; but the extra income is welcomed.
I bought the FY6600 for entertainment and learning value at a reasonable price.  Plus I thought the REAL contributors to this thread were doing amazing work and the few naysayers didn't have good enough arguments not to go down this rabbit hole.

[Miti]

Hey All,

I went through all (read most) of the pages but I couldn't figure out if Ver. 3.2 fixed the FP corruption issue. I apologise if I missed something.

[DaveR]

We're as sure as we can be that v3.2 fixed the problem.  No-one with a 3.2 device has ever reported any problems with it.

[DaveR]

Gandalf reminds me of someone who used to post in another forum of which I'm a member - nothing to do with electronics, but still just a hobbyist group.  We'd be discussing the intricacies and techniques of putting together another project, which would usually involve some reasonably small outlay for various parts and materials, and a fair amount of time for cutting and assembly (you know, the "hobby" part of the project), and this character would invariably chime in with something along the lines of, "Why are you wasting your time doing this when you can buy one for xxx dollars?" and would usually end by shouting "I REPEAT, HOW MUCH DO YOU VALUE YOUR TIME?".  The concept of a hobby and making something for pleasure, and retired people having lots of spare time to use as they see fit, all of which is free of charge, was completely alien to him and utterly beyond his comprehension - every waking hour (and probably sleeping time, too) had to have a price on it. 

He disappeared eventually, after everyone stopped replying to his posts.

[Miti]

We're as sure as we can be that v3.2 fixed the problem.  No-one with a 3.2 device has ever reported any problems with it.

Thanks,

Do you know if Banggood ships 3.2 or earlier?

[DaveR]

I got my v3.2 last December, so I'd be very surprised if there are any earlier versions still in circulation.  v3.2.1 is the latest firmware, and it's been around for a couple of months as well.  The chances of getting a 3,1 from any supplier must be miniscule by now.

[Miti]

Thanks Dave!

I just ordered one from BG. We'll see. Damn, another improvement project. 

[cybermaus]

...and retired people having lots of spare time to use as they see fit...

Well, this thread is good for my ego. I used to think of myself as one of the older guys, now I am starting to consider myself young... well, less old anyway.

But yes, its very simple, Gandalf_Sr is correct if your viewpoint is: "wasting hours tinkering" => this is not your device.
However, if you feel like "spending hours tinkering" => this may very well be a good device for you.

I do wish though he would answer the 200MHz/5ns jitter question for his Rigol.