Looking for design consultation MCI-JH24

[nickb]

The two head specs are so close it should just work, I would have thought. Clearly something is missing. The head inductance is specified at 1KHz. I wonder what it looks like at 105Khz?

There' always plan B Sony MCI JH-16 JH-24 Woelke 2" 24 Track Erase Head Assy Good Cond, Warranty JS | eBay

[nickb]

OK Mike, I gave this some more thought overnight. It's the low Q that's killing you. I wasn't paying too much attention to it before.

If the Q really is 0.77 (at 105KHZ) then the effective series resistance (winding and losses) is

2*pi*f*L/Q = 2 * 3.14 * 105000 * 0.0013 / 0.77 =1113 ohm for the Nortronic and
2 * 3.14 * 105000 * 0.0014 / 2.66 =347 ohm for the Woelke

The affects the ratio between the driving voltage from the transformer and that across the head (inductance plus resistance). For the Woelk that ratio it's 2.8 and the the Nortronic it's 1.2. To get more volts you have to change the transformer ratio (add more turns to the secondary) or add another ~1:2 ratio transformer on the output. If you can get spare transformers for this beast then I would certainly get one, measure it and see about adding the turns. If not, then it's time to add a transformer - you'd need about 4mH or more on the secondary side for a decent response at 105KHz.

[Westrek]

Thanks Nick.

Adding another transformer might just be the answer if I can find one small enough with the right specs.

Here's how they did it on the older machines when they used the Nortronics head. Which is the application I think the head that I'm trying to use is for. Both Bias and erase is 120K.

http://westendstudio.com/wp-content/...4/07/JH-16.pdf

[nickb]

Here's how they did it on the older machines when they used the Nortronics head. Which is the application I think the head that I'm trying to use is for. Both Bias and erase is 120K.

http://westendstudio.com/wp-content/...4/07/JH-16.pdf

Error 404 - not found

[Westrek]

Sorry, try this:

http://westendstudio.com/wp-content/...4/07/JH-16.pdf

[nickb]

Sorry, try this:

http://westendstudio.com/wp-content/...4/07/JH-16.pdf

Interesting. That gave me an idea. On the 24ch version, try connecting C3 to pin 6 instead of pin 5 of the existing transformer ( easier said than done, maybe install jumper too, huh?). The schematic hints that there might be an extra winding there. I think there is a good chance it will be just the right ratio.


(Edit) PS: You could always scope pin 5 and 6 to figure out what the ratio is.

[Westrek]

I did move it to pin 6 and it was a step in the right direction. Went from 38v to 45v.
As far as spec-ing a step up transformer, I just need to make sure the frequency response is adequate and try for a 2 to 1 ratio?
Most audio transformers have a 100k limit that I have seen. Do you have any suggestions on a proper transformer?

Mike

[nickb]

I did move it to pin 6 and it was a step in the right direction. Went from 38v to 45v.
As far as spec-ing a step up transformer, I just need to make sure the frequency response is adequate and try for a 2 to 1 ratio?
Most audio transformers have a 100k limit that I have seen. Do you have any suggestions on a proper transformer?

Mike

Bummer! I had my fingers crossed. I'll see what I can find but you might have to wind one yourself. In that case we would wind a replacement or the current transformer. Also, I can recommend someone to you who will make a new one for you based on looking at the current one.

At least we know the current transformer gets us half way there and so gives more flexibility.



It needs to handle about 4 Watts at 105Khz with a 1.1k load.

[nickb]

There are a couple of other ideas for things we can try but we have to be careful not to exceed the rating of the FETS. The rating that I think is of immediate concern is the max drain to source voltage of 60V.

We can adjust the T2 primary voltage in a couple of ways. One is to the adjust the slug of T2. As it gets nearer to resonance with C18 the voltage at the drains of the VN66's will increase. The problem is if it gets over 60V you risk damaging them. If we find that we are still nowhere near 60V then we can increase the gate drive voltage by increasing R36 and R39 slightly. With any luck it will be possible the raise the drive voltage high enough to get the erase current up. BTW, when you added FETs in parallel with the existing ones you increased the drive current by two.

So, put your scope on the drain of one of the FETs. If less than 60V peak adjust T2 slug to increase it. Using pin 6 of the transformer you were able to get to 45V on the head but you need 60V. So, you want to increase the T2 primary voltage by 60/45 = 33%. So long as this comes out less that 60V it will be possible to so with reasonable safely.

OF course we can also consider changing the FET to a higher voltage version.

[Westrek]

Thanks so much for your help! I'll be back on it tomorrow and report back.

Mike

[Westrek]

I constructed the coil loop and did the test. The frequency was 104.16K and the highest output I could get was .005v rms. That was with 1 track in record only.

[Westrek]

There is about 10v rms at the drain terminals of the vn66af in record and T2 peaked. Also, -5.6 dc at the drains.

[nickb]

There is about 10v rms at the drain terminals of the vn66af in record and T2 peaked. Also, -5.6 dc at the drains.

Thanks, Mike.

If the drains are really are -5.6VDC then the power dissipated in R41 would be over 3 watts (18-(-5.6))^2/180) and it is rated at just 3 watts so I have cause to suspect that measurement. Often meters get confused when measuring DC in the presence of AC. If you want to measure the DC voltage at the drains you'd do better to clip your meter to pin 2 of T2.

Can you also check the voltage at the other end of R41 too so I can work out the current used.

BTW, 10V rms is 28.3Vpp. Did you use the 'scope to do that? I wouldn't trust many meters at that sort of frequency. Based on that we need to get 60/45*28.3 = 37.8Vpp = 13.3Vrms.

It is starting to look like there is enough drive and we'll need to reduce R41 a little. There is always a risk that T2 will saturate with the higher primary voltage. I'm hoping it was designed with bit of margin.

If you could take those measurements then I'll know the next step.

Apologies for slow response. It appears that my instant notifications are not working and I can't seem to turn them on again.

EDIT: Duh! I forgot to allow for the drop on the output of IC1, Q11 and Q12 when calculating the power in R41. I estimate that drop at 2.7V making the power nearer (18-2.7+5.6)^2/180 = 2.42W. That's more believable. So, If you confirm the voltage at the top if R41 is a tad over 15V and the bottom of it is -5.6V than I think we can go ahead and replace R41 with a 120 ohm 5W ( 3W would do). Q13 and Q14 will run a bit warmer but should still be well within their limits. I think that will get the primary voltage around 13.5V rms (37.8vpp).

[Westrek]

Thanks Nick!

I'm going to be out of the loop for a day or so. I've done some other experiments and I will report back.

Mike

[Westrek]

Here's some measurements:

Stock (unmodified ) bias erase card:

T2 pin2 -5.2 vdc
Q13-14 Gate -17.6 vdc
" " Drain -5.12 vdc 30v p-p
" " Source -15.4 vdc
Erase output 38v rms

Modified card=Using pin 6 instead of 5 on T2 and R41 to 100 ohms.

Same test points:

-5.04 vdc
-16.24 vdc
-5.16 vdc 25v p-p
-14.3
Erase output 50v rms

I changed R36 and R39 both more and less from the 2K stock resistor. No real difference in erase output.