New Bipolar Boost TX and Front End

[KingJL]

@KingJL, I have been busy with restoring my computers and some test gear from power surge damage from blown line transformer.
As Murphy's law would have it, I normally have surge protection on all my work benches but I had set up a temporary area for these latest tests as I pulled out my ancient but trusty HP1630D which takes up a lot of bench space and used an unprotected power block.
Funny thing, The HP survived, 2 computers and some test gear did not.

Now back on deck and had a bit of time to test the latest Vivado project with Aux outputs.
Lucky for me I had the A7 and TX project safely tucked away on another bench and not connected

Two signals are missing
J3-7 = AUX_BSTB_S,
J3-8 = AUX_BSTB_E

[ATTACH]48575[/ATTACH]

All others look good. I will post better screenshots once I have a replacement GPIB interface card (casualty of the power surge)

Checked signals back at the A7 but both Aux_BoostB signals not there either.

The problem being that during the process of duplicating signals for the aux outputs, I duplicated BSTA_s and BSTA_e twice, eliminating the BSTB signals. This should take care of it. Also since the last posted project archive, I cleaned up some directories and files that Vivado was keeping around for modules that are no longer used.

Change log:
Change Log.txt

Current project archive is Vivado Bipolar_PI _Nextion(12-02-2019).zip

[Mdtoday]

The problem being that during the process of duplicating signals for the aux outputs, I duplicated BSTA_s and BSTA_e twice, eliminating the BSTB signals. This should take care of it. Also since the last posted project archive, I cleaned up some directories and files that Vivado was keeping around for modules that are no longer used.

Change log:
[ATTACH=CONFIG]48576[/ATTACH]

Current project archive is Vivado Bipolar_PI _Nextion(12-02-2019).zip

Thanks Jl, much appreciated. I will hook up and test again, then post screenshots later today

[KingJL]

Toroid Ferrite Core

I recevied various cores that I ordered to find one suitable for the Bipolar_PI. I had ordered 24mm and 22mm of J type ferrite (ui=5000), and 12.7mm of type W (ui=10000). The 24mm and 22 mm cores are too large to fit on the current footprint of the Bipolar_PI PCB. However the 12.7mm is the perfect size, both in diameter and height. Five turns on these cores calculates to 68uH which is the target range I was looking for. At a 4:1 secondary to primary ratio, a secondary of 20 turns = ~1100uH. With the new core, the secondary damping resistor ( R24, R30, R34, R39 ) should now be 810 ohm (vs. the current 1.5K). The charge holding resistor ( R25, R31, R35, R40 ) should now be 10 Meg ohm. The charge holding capacitor ( C29, C32, C35, C3 ) should now be 4n7 or 5n6.

I am going to wrap the primaries with AWG 22 tonight and check the inductance. I am waiting on a shipment of AWG 28 enameled magnet wire to wrap the secondaries.

For these 12.7mm cores we have an actual part number and a specification for the core and the ferrite material:
P/N ZW-41303-TC ( Magnetics )
type material: W ( 10000 ui )
outer diameter : 12.7mm
inner diameter : 8.14mm
height: 3.15mm
ui : 10,000

I think this more suitable than the cores of the modified chokes that we are now using as we can order a known P/N with known properties that we can depend on for repeatable results.

[KingJL]

Thanks Jl, much appreciated. I will hook up and test again, then post screenshots later today

The AUX_RX_BLK signal still is not right. It does not correctly follow the RX_BLK signal. Correction applied and new project archive will be posted shortly.

[KingJL]

The AUX_RX_BLK signal still is not right. It does not correctly follow the RX_BLK signal. Correction applied and new project archive will be posted shortly.

This version should have ALL of the AUX_* signals correctly functioning. This is what I get for not following through and running the Vivado simulator after a module is changed! All of the missing or incorrect signals would have been discovered by a Vivado simulation run of the TX-RX module. The AUX_RX_BLK was not following the RX_BLK signal after the initial TXA time.

Change Log:
Change Log.txt

Current project archive is Vivado Bipolar_PI _Nextion(12-03-2019).zip

[Altra]

I recevied various cores that I ordered to find one suitable for the Bipolar_PI. I had ordered 24mm and 22mm of J type ferrite (ui=5000), and 12.7mm of type W (ui=10000). The 24mm and 22 mm cores are too large to fit on the current footprint of the Bipolar_PI PCB. However the 12.7mm is the perfect size, both in diameter and height. Five turns on these cores calculates to 68uH which is the target range I was looking for. At a 4:1 secondary to primary ratio, a secondary of 20 turns = ~1100uH. With the new core, the secondary damping resistor ( R24, R30, R34, R39 ) should now be 810 ohm (vs. the current 1.5K). The charge holding resistor ( R25, R31, R35, R40 ) should now be 10 Meg ohm. The charge holding capacitor ( C29, C32, C35, C3 ) should now be 4n7 or 5n6.

I am going to wrap the primaries with AWG 22 tonight and check the inductance. I am waiting on a shipment of AWG 28 enameled magnet wire to wrap the secondaries.

For these 12.7mm cores we have an actual part number and a specification for the core and the ferrite material:
P/N ZW-41303-TC ( Magnetics )
type material: W ( 10000 ui )
outer diameter : 12.7mm
inner diameter : 8.14mm
height: 3.15mm
ui : 10,000

I think this more suitable than the cores of the modified chokes that we are now using as we can order a known P/N with known properties that we can depend on for repeatable results.

JL, thanks for sourcing a standardize core.

After following your's and MdToday's toroid experiments. I realized something was wrong with my thinking. The following is based on the cheap common mode chokes from ebay. First problem, I had my scope grounded to virtual ground. This was superimposing a 5v logic signal on to the floating GS signal. Once I put the scope ground on the source and the probe on the gate, I got a nice signal. Second problem, my mosfets have a max GS voltage of 20V, so the 4:1 ratio was too hot. Fortunately none of them were damaged. Out of four working Tx boards, one was perfect. It generated the highest current and voltage. I measured the GS voltage on each driver, all were about 10V. The toroids on the good pcb had a ratio of 2:1, 14t secondary and 7t primary. I had tried the 2:1 ratio multiple times with poor results. So I unsoldered all the toroids and measured the secondary all measured below 1900uH. Then I measured the 30 unused chokes that I have. These ranged from 1750uh to 2600uh. Out of these I selected 4 chokes with the lowest inductance. Winding 7t of 24 awg and leaving one of the original 14t windings. Next these were installed on one of poorly working Tx boards. The new toroids restored Tx operation to full output. Also the series resistors on the primary were changed to zero ohms. The core that were replaced had inductance's in the 2400uh range.

Not sure what to conclude? But am confident now that a GS voltage of 10v to 12v is more than ample. It appears that the higher inductance cores impede the 2us on/off pulse and degrade the TX output. Maybe some of this information will help. Thanks for sharing your work!

[KingJL]

... But am confident now that a GS voltage of 10v to 12v is more than ample.

I usually shoot for 15V-16V

It appears that the higher inductance cores impede the 2us on/off pulse and degrade the TX output. Maybe some of this information will help. Thanks for sharing your work!

Yes, if the primary inductance is too high, the short duration pulse (1.5 usec in my case) does not achieve optimum voltage (or more importantly, current). For a 1.5 usec triggering pulse, about 64-68 uH primary seems optimum. Also with a 64-68 uH primary, the resitors in the primary driving path can be increased back to 1 ohm, achieving a little faster charge current. With the common mode choke that were have been using, anything below 100 uH on the primary was difficult to get with any kind of accuracy. And with the extremely high permeability and such small dimensions, these common mode choke cores could have core saturation issues.

[Mdtoday]

I recevied various cores that I ordered to find one suitable for the Bipolar_PI. I had ordered 24mm and 22mm of J type ferrite (ui=5000), and 12.7mm of type W (ui=10000). The 24mm and 22 mm cores are too large to fit on the current footprint of the Bipolar_PI PCB. However the 12.7mm is the perfect size, both in diameter and height. Five turns on these cores calculates to 68uH which is the target range I was looking for. At a 4:1 secondary to primary ratio, a secondary of 20 turns = ~1100uH. With the new core, the secondary damping resistor ( R24, R30, R34, R39 ) should now be 810 ohm (vs. the current 1.5K). The charge holding resistor ( R25, R31, R35, R40 ) should now be 10 Meg ohm. The charge holding capacitor ( C29, C32, C35, C3 ) should now be 4n7 or 5n6.

I am going to wrap the primaries with AWG 22 tonight and check the inductance. I am waiting on a shipment of AWG 28 enameled magnet wire to wrap the secondaries.

For these 12.7mm cores we have an actual part number and a specification for the core and the ferrite material:
P/N ZW-41303-TC ( Magnetics )
type material: W ( 10000 ui )
outer diameter : 12.7mm
inner diameter : 8.14mm
height: 3.15mm
ui : 10,000

I think this more suitable than the cores of the modified chokes that we are now using as we can order a known P/N with known properties that we can depend on for repeatable results.

Excellent news, thanks JL.
I will check out local suppliers here for that part number.

[Mdtoday]

This version should have ALL of the AUX_* signals correctly functioning. This is what I get for not following through and running the Vivado simulator after a module is changed! All of the missing or incorrect signals would have been discovered by a Vivado simulation run of the TX-RX module. The AUX_RX_BLK was not following the RX_BLK signal after the initial TXA time.

Change Log:
[ATTACH=CONFIG]48583[/ATTACH]

Current project archive is Vivado Bipolar_PI _Nextion(12-03-2019).zip

Thanks JL, I didn't get to testing last night, should be back at it tonight. Ended up installing the new GPIB card and instrument drivers which did take me longer than expected.
Will load and test the latest version.

[Mdtoday]

I think this more suitable than the cores of the modified chokes that we are now using as we can order a known P/N with known properties that we can depend on for repeatable results.

I agree, the current cores tested show a lot of variation. With 18 turns secondary, 4 on the primary I was measuring 720-1200uH/ 24-48uH respectively with same winding method, so a better quality, specified component will help a great deal.

[Altra]

I agree, the current cores tested show a lot of variation. With 18 turns secondary, 4 on the primary I was measuring 720-1200uH/ 24-48uH respectively with same winding method, so a better quality, specified component will help a great deal.

Your average readings are about half of what I been measuring.

I bought 60 chokes from the same seller. Claimed 1.2mH, the ones received are mostly in the 2400uH range (14t). I assume my cores are completely different than the ones you guys have. I use an Extech LCR meter model 380193 which I trust. What can you expect for 10/$1.33 free shipping. Anyway I will be following to see if you two can get repeatable results with spec'd cores.

[Mdtoday]

Your average readings are about half of what I been measuring.

I bought 60 chokes from the same seller. Claimed 1.2mH, the ones received are mostly in the 2400uH range (14t). I assume my cores are completely different than the ones you guys have. I use an Extech LCR meter model 380193 which I trust. What can you expect for 10/$1.33 free shipping. Anyway I will be following to see if you two can get repeatable results with spec'd cores.

Yes, purchased mine from Ebay seller, they were worth a shot, think I paid more like $6 for 10 but tha'ts the Aus$. Measurements taken with Keysight U1733C and DER DE-5000 both within in a few uH of each other.
Trying to find a local supplier for the ones JL has posted, might have to purchase from the USA by looks of it.

[KingJL]

... Trying to find a local supplier for the ones JL has posted, might have to purchase from the USA by looks of it.

The linked "Magnetics" in post #528 shows an office in Hong Kong... surely there must be a source on that side of the planet!

[Mdtoday]

The linked "Magnetics" in post #528 shows an office in Hong Kong... surely there must be a source on that side of the planet!

Unfortunate but no stock in Asia and MOQ from Dexter Magnetics in the USA is 100.
There is an Ebay seller in the US as well, I'll keep searching, something will pop up.

[KingJL]

There is an Ebay seller in the US as well, I'll keep searching, something will pop up.

How many are you interested in? Ebay is where I got mine. Just wound 2 with primaries. The primaries were about 4 3/4 turns... My BM4070 measures at 70uH and 80uH... I will live with that variance.