| Electronics > Projects, Designs, and Technical Stuff |
| A "Poor Man's Adjustable Delay Line" |
| << < (3/4) > >> |
| tomato:
--- Quote from: dnessett on August 08, 2018, 11:01:35 pm ---Comparing this with Figure 8 of the previoius post, reveals significant smoothing (the Figure 8 trace is from 9 KHz - 20 MHz). The delay line is now probably useful up to around 85 MHz. So, the suggestion to use snap-on ferrite beads was a good one. --- End quote --- I don't think you can draw those conclusions from the graphs you presented: 1) The graphs don't cover the same frequency range, so a direct comparison can't be made. You have to compare the same measurement (with and without beads) if you want to draw a legitimate conclusion. For instance, the "ripple" in figure 8 has a periodicity of about 650 kHz. The resolution bandwidth in Figure 2 is 1 MHz, so there will be instrumental smoothing of the ripples as compared to Figure 8, which has a 100 kHz resolution bandwidth. 2) Figures 7 and 8 are essentially the same, yet Figure 7 has the coax completely switched out of the signal path. How can adding ferrite beads to the coax eliminate the ripple if the ripple exists even when the coax is switched out of the signal path? |
| dnessett:
--- Quote from: tomato on August 09, 2018, 02:37:43 am --- I don't think you can draw those conclusions from the graphs you presented: 1) The graphs don't cover the same frequency range, so a direct comparison can't be made. You have to compare the same measurement (with and without beads) if you want to draw a legitimate conclusion. For instance, the "ripple" in figure 8 has a periodicity of about 650 kHz. The resolution bandwidth in Figure 2 is 1 MHz, so there will be instrumental smoothing of the ripples as compared to Figure 8, which has a 100 kHz resolution bandwidth. --- End quote --- With the snap-on ferrites and all delay coaxes switched in. For comparison with Figure 8 in the previous post. --- Quote ---2) Figures 7 and 8 are essentially the same, yet Figure 7 has the coax completely switched out of the signal path. How can adding ferrite beads to the coax eliminate the ripple if the ripple exists even when the coax is switched out of the signal path? --- End quote --- At 10.0045 MHz in Figure 7 the signal is down -0.39 dB. At 10.0045 MHz in Figure 8 the signal is down -2.16 dB. |
| tomato:
--- Quote from: dnessett on August 09, 2018, 05:23:37 am ---With the snap-on ferrites and all delay coaxes switched in. For comparison with Figure 8 in the previous post. --- End quote --- That graph is much more useful. It looks good. --- Quote ---2) Figures 7 and 8 are essentially the same, yet Figure 7 has the coax completely switched out of the signal path. How can adding ferrite beads to the coax eliminate the ripple if the ripple exists even when the coax is switched out of the signal path? --- End quote --- --- Quote ---At 10.0045 MHz in Figure 7 the signal is down -0.39 dB. At 10.0045 MHz in Figure 8 the signal is down -2.16 dB. --- End quote --- The structure of the two graphs is essentially the same; the small difference in amplitude isn't very important. The question is, why are the ripples nearly identical when one of the graphs is taken with all of the coax is in the signal path, and the other graph is taken with all of the coax is eliminated from the signal path? |
| dnessett:
--- Quote from: tomato on August 09, 2018, 05:51:44 am ---The structure of the two graphs is essentially the same; the small difference in amplitude isn't very important. The question is, why are the ripples nearly identical when one of the graphs is taken with all of the coax is in the signal path, and the other graph is taken with all of the coax is eliminated from the signal path? --- End quote --- That is a reasonable question. However, my curiosity isn't raised enough to take off all of the snap-on ferrites and explore answers. The bottom line is the delay device is sufficient to work with 10 MHz signals and higher. I am revising the estimate of the upper limit of its usefulness, since I ran a TG trace using a RBW of 10 KHz and found structure below 85 MHz. Figure 1 shows the results. Figure 1 - So, the device as documented is probably good only up to 45 - 55 MHz. To push the design past this limit would require a redesign, for example one based on PC board trace transmission lines. Since such a design is significantly beyond my capabilities, I will leave it to others to pursue, if they are interested. |
| tomato:
--- Quote from: dnessett on August 09, 2018, 08:48:50 pm --- --- Quote from: tomato on August 09, 2018, 05:51:44 am ---The structure of the two graphs is essentially the same; the small difference in amplitude isn't very important. The question is, why are the ripples nearly identical when one of the graphs is taken with all of the coax is in the signal path, and the other graph is taken with all of the coax is eliminated from the signal path? --- End quote --- That is a reasonable question. However, my curiosity isn't raised enough to take off all of the snap-on ferrites and explore answers. The bottom line is the delay device is sufficient to work with 10 MHz signals and higher. --- End quote --- Really, not at all curious? The two graphs are essentially identical, i.e. every little feature in Figure 7 appears in Figure 8. That just screams out something is wrong with one or both of the measurements. |
| Navigation |
| Message Index |
| Next page |
| Previous page |