[sdantonio]

Hi Mariss,

I was looking in another post and I saw the G540 (4 axis, 3A) mentioned. If it's still in the design stage any possibility it can be kicked up to a 5 or 6 amp rating? Or possibly even putting together a G541 with the typical 7A, 80VDC rating that my other Gecko's have.

Steven

[alipavsky]

I've been contemplating power supplies for the G250's. Would there be an effective, inexpensive, and recommended way to attenuate the primary side of a standard step-down transformer to provide the required ~36V on the secondary side?

My thinking is that step-up/down transformers are typically much less expensive and more readily available than 36V transformers or switched power supplies. They're just a little too hot, right out of the box.

- Andrey -

[Mariss Freimanis]

Sdantonio,

Not a snowball's chance in Hades. Cost of components and size tracks a drive's power output. The G250 is meant to service 3A motors that presently underutilize our 7A, 80V drives. $114 is not a bad price for a 7A, 80V drive. It is less of a bargain if you have a 3A motor and the G250 is meant to remedy that.

Mariss

[Mariss Freimanis]

Andrey,

A 24 VAC secondary transformer gives 34VDC after rectification and filtering.

Mariss

[alipavsky]

Andrey,

A 24 VAC secondary transformer gives 34VDC after rectification and filtering.

Mariss

Mariss,
I was refering to conditioning 36VAC to 50VDC. The potential 1/3 speed gain is too tempting to pass up.

Incidently, do the G250's care if the power comes from a transformer or a switching supply?

Andrey

[Mariss Freimanis]

Makes no difference what the DC source is.

Mariss

[alipavsky]

So back to the original question, what's the verdict on attenuating the primanry of a standard step-down transformer to produce 36VAC instead of 55VAC?

[KYYADA]

If I understand your question the wire sizes in the transformer will be different. The primary will have small wire which can handle the reduced amperage because of the higher voltage. The secondary will have bigger wire to handle the decreased voltage but higher amperage. You will acheive the difference in voltage because of the ratio, but the amp rating will be different. If it has a fuse on the primary side that would be a good indicator of the new rating.

[KYYADA]

Thinking about that a little more if you fliped the ratio from say 3:1 to 1:3 you would multiply the 110 by a factor of 3. I think! If you have 110 turns on the primary and 110 Volts input then for every turn on the secondary you would get 1 volt.

[alipavsky]

If I understand your question the wire sizes in the transformer will be different. The primary will have small wire which can handle the reduced amperage because of the higher voltage. The secondary will have bigger wire to handle the decreased voltage but higher amperage. You will acheive the difference in voltage because of the ratio, but the amp rating will be different. If it has a fuse on the primary side that would be a good indicator of the new rating.

Thinking about that a little more if you fliped the ratio from say 3:1 to 1:3 you would multiply the 110 by a factor of 3. I think! If you have 110 turns on the primary and 110 Volts input then for every turn on the secondary you would get 1 volt.

No, that's not quite what I'm talking about. This is all about providing the 50VDC max to the G250s. A standard step-down transformer will provide ~55VAC, conditioned to ~72VDC.

If that primary coil is attenuated so that ~36VAC is induced on the secondary, that will provide the ~50VDC needed.

Andrey

[Mariss Freimanis]

Hi,

I thought I would fill everyone in on the current status of the G251 (G250 is the 'dirt-cheap' version, G251 is the 'cheap' version with mounting plate and screw-terminal connector block).

The G251 REV2 prototype works properly. It has been running at 3.5A and 60VDC all weekend since Friday afternoon. It is sitting on the lab bench with no heasinking on its 1.6" by 1.6" mounting plate. The plate reads 68C while the ambient is 27C (it's a warm weekend here).

The G251 has been run-up to a 330kHz step pulse frequency (9,900 RPM on the test motor). It functions properly from 13.5VDC to 60VDC. The rated voltage of 15VDC to 50VDC seems to be a reliable and safe rating. Current has been tested from 300mA (NEMA-17) to 3.5A (NEMA-34) and all scope traces show normal and clean waveforms. The motors run very smoothly.

Mid-band resonance compensation has been tested at both ends of the spectrum. Damping is very pronounced and effective indicating a good phase margin.

Microstep to full-step morphing was changed from on/off to a proportional response like in the G203V. This required the addition of a few components and it seems doing that was worth the trouble.

In effect, the G251 meets all design goals and exceeds a few, heat dissipation being much less than originally expected.

This weekend I'm incorporating the morphing circuit changes on the pcb layout. The layout gets proofed by someone else here (never ever double-check your own work; have someone else who is expecting to find flaws do that). Tuesday the gerber files go out for pcb panel fabrication and the first production run should be the 3rd week of April.

The thumbnail shows the assembled G251 before it was put through instrumentation and testing. The production units will look just like this except the plate will be hard-anodized and black in color.

Mariss

[Kipper]

We will have more contests coming up before the geckos are released into the wild, so be sure to check back.

Video of it going into the "Mr Wild" will be on Youtube?

[skmetal7]

so how does microstepping/morphing work? what would i set in mach3 for steps/inch if i have a .2" lead and a 200 steps/inch stepper? won't steps/inch change? and how does the drive handle morphing, will it go to 1/4 step or 1/2 step at a faster speed?

good to hear the progress report!

[eman5oh]

so how does microstepping/morphing work? what would i set in mach3 for steps/inch if i have a .2" lead and a 200 steps/inch stepper? won't steps/inch change? and how does the drive handle morphing, will it go to 1/4 step or 1/2 step at a faster speed?

good to hear the progress report!

A .2 lead means 5 turns per inch, that times 200 steps per turn will give you 1000 full steps per inch. With a 10th microstep drive you have 10000 steps per inch for your mach config. As far as the morphing, the drive switches from a sine like current wave form to a square current wave form when the motor reaches a set speed. This all happens automatically inside of the drive. You do not need to do any thing special with mach3 or other controllers for that matter. You conversion will remain a constant 10000 steps per inch. Eric

[tooManyHobbies]

Mariss,

Am I correct in saying that the G540 will use the G250, not the G251?

It looks like the heatsink of the 251 would get in the way of its neighbor and that the 540 would utilize the frame itself for heatsinking.

Just curious.

tom

[Ch_Irawan]

The layout gets proofed by someone else here (never ever double-check your own work; have someone else who is expecting to find flaws do that).

Is this about me Mariss? If yes then I need to make it clear, I did not expecting to find any flaws on the design; no other interest except the spirit of competition and curiosity on electronics (learn something). I am curious to anything related to electronics, for example I always open brand new mobile phone before using it. It will be a different story if you ask me to do so.

Best Regards, Ichan.

PS: You have not answer my questions on the group (???)

[microcarl]

I use to have a boss back in the 1980's - the owner of a small company, who was also a good friend. His name was John Kerr. Anyway, I would frequently go to him with ideas. If John liked a particular idea, he would dedicate my time to it exclusively.

In every case, John would pop in my office at least two or three times a day and ask "Is it possible to do this?" or, "Is it possible to do that?"

Well, when the idea was first presented, it was usually sound and practical. But by the time the project was well along it's way, and John had added all of the extra periphery, it had so many "Bells & whistles" that it was either too difficult to manufacture or the cost that the customer would have to bare was simply too high. Since that time, I have termed "Feature Creep" as "Kerr-izm", because John Kerr was the master of "Feature Creep!"

So, what I see here is a similar situation... The engineers of GeckoDrive have come up with a series of fantastic ideas. But if they design in every request that has been presented by everyone in this thread, you all won't be able to afford any of these great new products that will soon be made available to us.

So, I ask you, does what I've said make any sense at all?

[Mariss Freimanis]

Ch_Irawan,

1) No, it's not about you. It's about how to cancel one's own ego when it comes to evaluating your own work. You design something and what comes with it is a built-in prejudice that your work is correct. That prejudice is natural and one cannot ever depend on being able to successfully overcome it. For that reason, others sit in judgment during a design review and they are encouraged to assume the work is flawed. During a pcb layout proof, they undo your work literally a trace at a time. Remove a trace from the artwork, pencil it in in the schematic. The proof is finished when the layout is missing all traces and every connection in the schematic is penciled in.

2) I know, I promised but the whole weekend has been absorbed in pcb layout changes. I'll try to answer it quickly here, in more depth on the Yahoo group tomorrow.

Anytime both MOSFETs in a half-bridge get killed is a sign of a short-circuit somewhere. The top MOSFET gets killed by the original short to ground and it goes to zero Ohms. When the short gets released, the bottom MOSFET is working into a zero-Ohm load in its drain circuit and it fails next. The resulting failed MOSFET pair crowbar the power supply.

All 4 half-bridges in the drive are identical. The odds of any one half-bridge failing are 1 in 4. You said you have 3 drives with failed phase 'C' half-bridges. The odds are 64 to 1 against it being a random occurrence. Something is going south with phase 'C' in your mechanisms.

Mariss

[Mariss Freimanis]

Feature creep is either due to an undisciplined designer or more often a designer that is at the mercy of a committee. Any good new idea initially coalesces into a clear, simple and compact essence of what is to be achieved. When the idea is explained, others agree it is a good new idea but it could even be better if this or that feature is added. Left unchecked, everything but the kitchen sink gets thrown in and the result is an unwieldy Frankenstein mutation of what was originally intended. Reminds of Melanie's "Look What They've Done to My Song, Ma".

Mariss

[microcarl]

Mariss,

I hope you realize that my "Feature Creep" comments weren't aimed at you. It's just with all of the request for features, the whole thing reminded me of what I went through with my friend.

As a fan of the GeckoDrive, I think your recent product presentations are excellent. I'm pleased to see that you do have the discipline to control the "Feature Creep."

And yes! My friend John was quite undisciplined. But he was the owner of the company, who was providing me with a paycheck. So, "Feature Creep" was the way it had to be. That, or find another employer.

You certainly make GeckoDrive a great company and you also have some great products and great ideas!

Thanks!