The AM Forum
March 28, 2024, 04:59:46 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Gate driver ICs  (Read 24189 times)
0 Members and 1 Guest are viewing this topic.
KQ6F
Member

Offline Offline

Posts: 124


« on: October 27, 2016, 03:41:39 PM »

I've been using TC4452VAT gate drivers with my SiC RF deck and have become very dissatisfied.  They seem to be fragile....I've blown up 2 or 3 of them under mysterious circumstances.  They are adequately heatsinked, +15v supply properly bypassed, about 1000 pf load.  Operating 75M.

I notice that others are using the IXDD614CI driver and am wondering how that's been working out.  Any unexplained failures?

Rod
Logged
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #1 on: October 27, 2016, 07:10:50 PM »

I've been using that series (the IXDD414 and the IXDD614) for several years, and they work very well.  Pretty easy to use, too.  And not too expensive. Only good up to 40 meters.
Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #2 on: October 30, 2016, 01:00:13 AM »

I have an IXDD614 driving an IRF510 on 20 meters CW.  It's not happy at all, but I manage to get 5 watts out of it.  I increased the voltage on the FET to 48 volts and fire shot out of one of the two and the other flew across the room.  I rebuilt it and leave everything at 13.8 vdc.

My farthest contact thus far with it has been Alaska.

Jon


* IMG_0988.JPG (2349.53 KB, 4032x3024 - viewed 1305 times.)

* IMG_0990.JPG (2748.43 KB, 3024x4032 - viewed 1082 times.)
Logged

It’s not just values, it’s business.
steve_qix
Contributing
Member
*
Offline Offline

Posts: 2599


Bap!


WWW
« Reply #3 on: October 31, 2016, 10:00:11 AM »

Yeah, the IXDD series isn't made to go that high in frequency.  Even 40 meters is a bit of a stretch.

Best bet at those frequencies and above may be analog drive.  I've experimented with analog drive on 10 meters a bit.  It is workable.
Logged

High Power, Broadcast Audio and Low Cost?  Check out the class E web site at: http://www.classeradio.org
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #4 on: October 31, 2016, 08:00:44 PM »

I use analog drive on my ten meter deck.  Works fine.

The CB scene is using a couple resistors and a diode along with the IRF520 to get decent efficiency at 11 meters. 

http://www.cbtricks.com/miscellaneous/fet_papers/mosfet_experiment.htm

The information there is useful regardless of frequency.   I am contemplating using one of those circuits on my 500 watt pep ten meter deck to improve efficiency. 

--Shane
KD6VXI
Logged
KA6MOK
Member

Offline Offline

Posts: 22


« Reply #5 on: November 05, 2016, 06:20:20 PM »

Hey all...

Haven't been back here for a while,  work and etc...

Wanted to pass along what I'd learned in my own searching on the subject.  Hope it's helpful.
I was wondering about Nigel's problems with (apparent) failures of the IXDD630 parts, and I actually got ahold of an applications engineer at IXYS Colorado to ask about that.  Even though the IXDD parts aren't theirs, (different divison) he had some suggestions. 

Main thing was,  the faster and higher current the driver is, the more critical the layout of the power and ground and the type and amount of the bypassing parts get.   This is somewhat of an obvious point,  but DIY'ers like us may be underestimating how big a deal it is.  He said he's seen a lot of pictures of scary long, point to point wiring, from RF engineers prototyping their designs and calling them with problems...

He also suggested that due to non optimal layout and bypassing,  it could very well be possible to have ground bounce that could cause parts to blow themselves up.   Now, in the other thread, it turns out that Nigel didnt actually lose parts,  but they seem to have an unexpectedly high undervoltage lockout.  I'm wondering if the same effect can cause them to not correctly do the undervoltage lockout, and explain the other erratic behavior.

Other things I've found of possible interest,  there's a couple of nice app notes on the IXYS Colorado site, that cover layout and grounding of the fast driver IC's:

An evaluation board for the uber chip, the IXRFD630:
http://www.ixyscolorado.com/index.php/ixys-rf/mosfets-drivers-diodes-power-modules/evaluation-boards/item/dvrfd630-631

And an evaluation board for that driver and one of their similarly packaged FETS, in a 1KW, 13.5Mhz class E amp. (Oooh, hope for 20M!!) This also goes into what's good for optimal driver chip conditions:
http://www.ixyscolorado.com/index.php/ixys-rf/mosfets-drivers-diodes-power-modules/evaluation-boards/item/prf-1150

Now, of course this is for the ultra high end driver IC which is much faster than the IXDD630,  but it gives you the idea of what is important.  The lower end ones may OK with less perfect layout and bypassing.  But since the IXDD630 is about twice as fast as the IXDD614,  it might need more attention.  And, doing the multiple sized and very short or SMT bypass parts will help some with the slower parts, too.

Lastly, looks like IXYS Colorado did a facelift to their website, not much new, but apparrently there's a few new ZX series FETs, some that look very interesting for class E.  Unfortunatly, they seem to be in the >$40 range...
Logged
VE3ELQ
Contributing
Member
*
Offline Offline

Posts: 258


« Reply #6 on: November 05, 2016, 07:54:22 PM »

Jon,
Thanks for your research and the post.  The comments about robust bypassing and short leads are well received, indeed I did come to understand that and endeavored to provide such bypassing with 3 parallel caps as close as possible to the Vcc pin and ground.  I even ran a separate ground wire to a lug right on the chip all to no avail.  These drivers just need 20 volts to function to overcome the UVL, then they work.  Every time I install a driver I keep thinking why dont these things come in a flat pack with input stuff on one side and output on the other to keep things short and there in the link you provided for the IXRFD630 shows just such a device, but they cost $31 each up here and no UVL spec mentioned.

My project using IXDD614CI drivers has now developed into a band switched, single deck, 3 band 350W transmitter for 160 80 and 40 meters. The deck is working exceptionally well providing virtually UN-measurably high efficiency on 160 and 80 meters and 93 % on 40M all class D. At a later date I may try 20M but will probably use the Microsemi DRF1300 which includes the FETs and drivers.

73s  Nigel
Logged

A "Tube Free" zone.
KA6MOK
Member

Offline Offline

Posts: 22


« Reply #7 on: November 06, 2016, 04:31:45 AM »

Hmmm... it's strange that they would need so much voltage to run.  The spec is supposed to be nominally 12.5V, with a range of 10-13.5V.  You almost never see parts not matching their data sheets,  so to see a max 13.5V threhsold take 20V to work... seems to say there's something else going on.  Think that's worth talking about to an applications guy again.   I need to find the other IXYS branch that used to be Claire,  who makes the IXDxxxx parts.  Hope you didn't get a batch of Chinese counterfits!

Good to hear it's solid and efficient on 40 with the 614's... that's a great result!  20 would probably need the higher speed.

I have noticed that IXYS Colorado has some integrated parts like the Microsemi's, such as this thing: http://www.ixyscolorado.com/index.php/ixys-rf/mosfets-drivers-diodes-power-modules/rf-power-modules/item/ixz631df12n100 Looks like the holy grail, probably could run happily on 10M...
I haven't priced them to see how they compare. Likely $$.  Probably $$$.   Actually, I haven't found anyone selling them yet to find out...

Oh, and I forgot to point out earlier,  the 1KW 13.5Mhz amp article has a nice duty cycle adjust circuit using a D flip flop and a few parts... worth checking out.  Should have a good adjustment range, and using the suggessted AC part it won't be slowing things down much, till somewhere beyond 20M...

Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #8 on: November 06, 2016, 05:12:07 PM »

Does anyone know who stocks the IXDD630's with the TO-220 package? 

I've looked at Mouser, DigiKey and Newark and they stock a different package type.  I found one eBay listing from Germany selling a single chip for over $15. 

Jon
Logged

It’s not just values, it’s business.
VE3ELQ
Contributing
Member
*
Offline Offline

Posts: 258


« Reply #9 on: November 06, 2016, 08:41:13 PM »

Here ya go:
http://www.digikey.com/product-search/en?keywords=%20%09CLA374-ND%20
Good luck with em.
73s  Nigel
Logged

A "Tube Free" zone.
KA6MOK
Member

Offline Offline

Posts: 22


« Reply #10 on: November 06, 2016, 11:20:22 PM »

Here's a bit better price...  Arrow has them for $5.35,  both the inverting and non inverting.
https://www.arrow.com/en/products/ixdi630ci/ixys?utm_source=octopart&utm_medium=buynow&utm_campaign=octopart

I haven't ordered from them in a while... don't know if they are too much more of a hassle than Digikey.

Seems nobody's selling small quantities of the M suffix (the 9V undervoltage) parts in TO-220.  They are around in that TO-263 package, also at Arrow (and I think Digikey and Mouser).

A thought, if they turn out to need more serious layout and bypassing,  i.e. putting them on a patch of PCB,  then one might end up prefering the SMT style package over the TO-220.  Of course the tradeoff is they're a bit more difficult to replace when they fly across the room... Wink

Jon
Logged
ka1tdq
Member

Offline Offline

Posts: 1509


Red part turned in for a refund.


« Reply #11 on: November 07, 2016, 06:07:52 AM »

Cool!  The supply voltage will need to be a minimum of 12.5vdc, so 13.8 would probably work out pretty good. 

I just bought this from eBay.  It's the dial faceplate from a 1927 era RCA radio.  I'm sure I'll build another class E rig sometime, so I'll use this around the front of the tuning variable capacitor.  For the hole where the dial goes, I'll invert an LED bar graph to measure drain current.  The top hole would be a bright LED to indicate driver bus voltage. 

I just got some Brasso to clean it up.  Should look pretty neat!

Jon


* IMG_1042.PNG (3347.17 KB, 1242x2208 - viewed 1012 times.)
Logged

It’s not just values, it’s business.
VE3ELQ
Contributing
Member
*
Offline Offline

Posts: 258


« Reply #12 on: December 04, 2016, 11:38:40 AM »

Tested the NCP81074 drivers running at 9V with C3M0280090D SIC FETs at 14.3mhz and 7.4mhz, one driver per FET.  The 20m gate waveforms looked pretty good and the 2 FETs made 120W RF but only at 85% efficiency in class E. The deck got pretty warm. These FETs are not quite fast enough for efficient 20m operation but they do work up there.
On 40m the FETs worked super well with un-measurably high efficiency, deck barely warm at 160W output.
These are very good little drivers on a small PCB only slightly larger than an IXDD614 but 8 times faster and inexpensive.  For parallel FETs one driver per FET would be ideal.  Just some info. PM or Email if you want the driver PCB files.

73s  Nigel
Logged

A "Tube Free" zone.
kf2vm
Member

Offline Offline

Posts: 8


« Reply #13 on: December 10, 2016, 06:36:38 PM »

Hi Everyone,

 I have had a great deal of success with both the IXRFD630 and the IXRFD631. I use them through-out my Shortwave Broadcast line.
They are very rugged,...and reliable.

 The IXRFD630 is fine for 10 MHz. and below. They do exhibit ground bounce problems above about 12 MHz. Vdd decoupling gets very problematic, as the Vdd tabs are opposite the ground tabs.

 The ground bounce moves the device threshold voltage around,..causing the pulse width of the R.F. waveform to jitter starting around 13 Mhz.
   I experienced this when prototyping R.F.P.A. modules for a 10 k.w., 10 to 20 MHz transmitter.
 
  Ixys came out with the IXRFD631 while I was in development.

 This device has a Kelvin Ground right next to the input pin. This allowed me to use a common mode choke on the input and eliminated the issue.

 I'm using a pair of IXRFD631 to drive a pair of IXZ318N50 in each R.F.P.A. "sub-module"......(push-pull). I then combine 8 "sub-modules" to make a 1.2 K.W. chassis.
 
 Then combining 10 chassis to make 10 K.W.

 They work great up to 20.250 MHz. That's as high as I've gone so far.

 They can be rather pricey for amateur service,...as well as difficult to get. Everyone in the Solid State R.F. industry uses them,....because they are just so darned reliable.

 However,..if Ixys is giving you a long lead time,.......you can be pretty confident it is because Digikey has just made a large purchase from them. Digikey has been my fallback when Ixys is out.

 In regards to the "composite" driver/M.O.S.F.E.T. devices,........I've had some problems meeting linearity requirements when using them. (Less than 1% distortion / -46dB I.M.D. for D.R.M. requirements in shortwave broadcast.)

 Honestly,...I don't know why this is. I didn't have the time to explore the issue. Additionally,...it may be a non-issue in typical Amateur Service.

 There's also the possibility I was missing something,....so I don't wish to discourage anyone from trying them.

 I've been through the ringer with Gate Drivers. The IXRFD630/631 has been the best I've seen.

                                                                      Best Regards
                                                                         Jamie


Logged
Tom WA3KLR
Contributing
Member
*
Offline Offline

Posts: 2120



« Reply #14 on: December 11, 2016, 09:03:35 AM »

Hey Jamie,

It's terrific to "hear" from you!  Merry Christmas to you.

Tom WA3KLR

Logged

73 de Tom WA3KLR  AMI # 77   Amplitude Modulation - a force Now and for the Future!
kf2vm
Member

Offline Offline

Posts: 8


« Reply #15 on: December 11, 2016, 08:07:07 PM »

Merry Christmas to you also Tom!!!

I don't want to derail the thread.

I will be back home for Christmas on December 17th. Working on a project here in Fresno.

                                                    Jamie
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #16 on: December 11, 2016, 09:35:19 PM »

Howdie neighbor,  Bakersfield checking in.

Have hf capabilities?   And great info on your last post.   I've accomplished sinewave drive class e on 10, wanted to find drivers that would work there.....

Glad to see things are getting closer!

--Shane
KD6VXI
Logged
kf2vm
Member

Offline Offline

Posts: 8


« Reply #17 on: January 03, 2017, 08:11:15 PM »

Howdy Bakersfield,

   Y'know,......give the IXRFD631 a try on 10 meters!!! Especially if whatever your driving has relatively low Gate Capacitance.
  I'm sure I don't have to tell you how critical de-coupling is,...and all of the boilerplate layout precautions.

  In theory,....these supposedly work to 45 Mc/s. But,.....we all know how all the variables affect that "specification".

  Came home for the Holidays,.....but heading back to Fresno on Thursday January 5th.

  No H.F. capabilities as I'm staying in a Hotel Room. I do listen via WebSDR. Every now and then I go to Mike, K6MYC of M2 Antenna Systems and play radio.
 
 Also Ken,...WA6OIB.

 Great job swinging those gates at 28 Mc/s!! I might give it a try at 25 to 26 Mc/s,....as there is a D.R.M. broadcast band in Europe on,....yes,.....11 METERS!! (Well,....a bit below.)

                                                                            73 DE KF2VM
                                                                              (Jamie)

Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.079 seconds with 19 queries.